CN114343574A - Sleep quality evaluation method, client, server and storage medium - Google Patents
Sleep quality evaluation method, client, server and storage medium Download PDFInfo
- Publication number
- CN114343574A CN114343574A CN202111602630.2A CN202111602630A CN114343574A CN 114343574 A CN114343574 A CN 114343574A CN 202111602630 A CN202111602630 A CN 202111602630A CN 114343574 A CN114343574 A CN 114343574A
- Authority
- CN
- China
- Prior art keywords
- model
- sample
- sleep
- parameters
- training
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 230000003860 sleep quality Effects 0.000 title claims abstract description 154
- 238000013441 quality evaluation Methods 0.000 title claims abstract description 82
- 238000003860 storage Methods 0.000 title claims abstract description 14
- 238000000034 method Methods 0.000 title claims description 44
- 238000012549 training Methods 0.000 claims abstract description 173
- 230000007958 sleep Effects 0.000 claims abstract description 165
- 238000001303 quality assessment method Methods 0.000 claims abstract description 65
- 230000015654 memory Effects 0.000 claims description 42
- 230000007613 environmental effect Effects 0.000 claims description 34
- 238000012360 testing method Methods 0.000 claims description 18
- 238000004590 computer program Methods 0.000 claims description 15
- 238000004891 communication Methods 0.000 claims description 13
- 238000007637 random forest analysis Methods 0.000 claims description 9
- 238000011156 evaluation Methods 0.000 claims description 6
- 238000013473 artificial intelligence Methods 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 19
- 238000012545 processing Methods 0.000 description 13
- 230000001133 acceleration Effects 0.000 description 10
- 238000012544 monitoring process Methods 0.000 description 8
- 238000003066 decision tree Methods 0.000 description 7
- 238000010801 machine learning Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 4
- 208000019116 sleep disease Diseases 0.000 description 3
- 208000020685 sleep-wake disease Diseases 0.000 description 3
- 230000000903 blocking effect Effects 0.000 description 2
- 238000004140 cleaning Methods 0.000 description 2
- 238000013500 data storage Methods 0.000 description 2
- 230000003203 everyday effect Effects 0.000 description 2
- 230000005484 gravity Effects 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 230000004617 sleep duration Effects 0.000 description 2
- 241000282326 Felis catus Species 0.000 description 1
- 206010062519 Poor quality sleep Diseases 0.000 description 1
- 208000013738 Sleep Initiation and Maintenance disease Diseases 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000002354 daily effect Effects 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 206010022437 insomnia Diseases 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 230000004620 sleep latency Effects 0.000 description 1
- 230000004622 sleep time Effects 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
Images
Landscapes
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The embodiment of the application discloses a sleep quality assessment method, a client, a server and a storage medium, which relate to the technical field of artificial intelligence and comprise the following steps: sending a model obtaining instruction to a server to obtain a first model from the server based on the model obtaining instruction; training a first model by using sample environment parameters corresponding to the sample object and the sample sleep label to obtain a sleep quality evaluation model; the sample environment parameter is the environment parameter of the environment where the sample object is located; the sample sleep label is a sleep label corresponding to the sample environment parameters; and under the condition of receiving the to-be-tested environment parameters corresponding to the to-be-tested object, inputting the to-be-tested environment parameters into the sleep quality evaluation model to obtain a sleep quality evaluation result corresponding to the to-be-tested object.
Description
Technical Field
The application relates to the technical field of artificial intelligence, in particular to a sleep quality assessment method, a client, a server and a storage medium.
Background
The quality of sleep has an important influence on the physical health of contemporary people. Due to the rapid development of the current internet, more and more night cats on the internet are emerged and are addicted to the world of the network, and people enjoy the audio-visual impact of the fast traffic of the network and are accompanied with the trouble of irregular sleep and insomnia. Therefore, people need to monitor sleep and improve the body health index of people.
In the prior art, a large amount of user sleep monitoring data are collected and stored in a remote cloud center, and the user sleep monitoring data are analyzed to determine the sleep quality of a user.
Disclosure of Invention
In order to solve the above technical problems, embodiments of the present application are expected to provide a sleep quality assessment method, a client, a server, and a storage medium, which can improve the security during sleep quality assessment.
The technical scheme of the application is realized as follows:
the embodiment of the application provides a sleep quality evaluation method, which is applied to a client and comprises the following steps:
sending a model obtaining instruction to a server to obtain a first model from the server based on the model obtaining instruction;
training the first model by using sample environment parameters and sample sleep labels corresponding to the sample objects to obtain a sleep quality evaluation model; the sample environment parameter is the environment parameter of the environment where the sample object is located; the sample sleep label is a sleep label corresponding to the sample environment parameter;
and under the condition of receiving the to-be-tested environment parameters corresponding to the to-be-tested object, inputting the to-be-tested environment parameters into the sleep quality evaluation model to obtain a sleep quality evaluation result corresponding to the to-be-tested object.
The embodiment of the present application further provides a sleep quality assessment method, applied to a server, including:
under the condition of receiving a model acquisition instruction transmitted by a client, acquiring a first model corresponding to the model acquisition instruction;
and sending the first model to the client, so that the client trains the first model based on the sample environment parameters and the sample sleep labels at the client to obtain the sleep quality evaluation model.
An embodiment of the present application provides a client, where the client includes:
the first sending unit is used for sending a model acquisition instruction to the server;
a first obtaining unit, configured to obtain a first model from the server based on the model obtaining instruction;
the training unit is used for training the first model by using the sample environment parameters and the sample sleep labels corresponding to the sample objects to obtain a sleep quality evaluation model; the sample environment parameter is the environment parameter of the environment where the sample object is located; the sample sleep label is a sleep label corresponding to the sample environment parameter;
and the input unit is used for inputting the environmental parameters to be tested into the sleep quality evaluation model under the condition of receiving the environmental parameters to be tested corresponding to the object to be tested, so as to obtain the sleep quality evaluation result corresponding to the object to be tested.
An embodiment of the present application provides a server, where the server includes:
the second acquisition unit is used for acquiring a first model corresponding to a model acquisition instruction under the condition of receiving the model acquisition instruction transmitted by the client;
and the second sending unit is used for sending the first model to the client so that the client trains the first model based on the sample environment parameters and the sample sleep labels at the client to obtain the sleep quality evaluation model.
An embodiment of the present application further provides a client, where the client includes:
the sleep quality assessment method comprises a first memory, a first processor and a first communication bus, wherein the first memory is communicated with the first processor through the first communication bus, the first memory stores a program of sleep quality assessment executable by the first processor, and when the program of sleep quality assessment is executed, the sleep quality assessment method applied to a client side is executed through the first processor.
An embodiment of the present application further provides a server, where the server includes:
the sleep quality assessment method comprises a second memory, a second processor and a second communication bus, wherein the second memory is communicated with the second processor through the second communication bus, the second memory stores a program of sleep quality assessment executable by the second processor, and when the program of sleep quality assessment is executed, the sleep quality assessment method applied to the server is executed through the second processor.
The embodiment of the application provides a storage medium, on which a computer program is stored, and is applied to a server and a client, wherein the computer program is executed by a first processor to implement the sleep quality assessment method applied to the client; the computer program, when executed by the second processor, implements a sleep quality assessment method applied in the server.
The embodiment of the application provides a sleep quality evaluation method, a client, a server and a storage medium, wherein the sleep quality evaluation method comprises the following steps: sending a model obtaining instruction to a server to obtain a first model from the server based on the model obtaining instruction; training a first model by using sample environment parameters corresponding to the sample object and the sample sleep label to obtain a sleep quality evaluation model; the sample environment parameter is the environment parameter of the environment where the sample object is located; the sample sleep label is a sleep label corresponding to the sample environment parameters; and under the condition of receiving the to-be-tested environment parameters corresponding to the to-be-tested object, inputting the to-be-tested environment parameters into the sleep quality evaluation model to obtain a sleep quality evaluation result corresponding to the to-be-tested object. By adopting the method, the client acquires the first model from the server by utilizing the sample environment parameters corresponding to the sample object and the sample sleep label training, so as to obtain the sleep quality evaluation model, the sample environment parameters of the sample object do not need to be sent to the server, the risk of leakage of the sample environment parameters of the sample object or the to-be-tested environment parameters corresponding to the to-be-tested object does not exist, and the safety during sleep quality evaluation is improved.
Drawings
Fig. 1 is a prior art sleep quality assessment network model training diagram based on conventional machine learning according to an embodiment of the present disclosure;
fig. 2 is a flowchart of a sleep quality assessment method according to an embodiment of the present disclosure;
FIG. 3 is a schematic diagram of an exemplary data collection module provided by an embodiment of the present application;
FIG. 4 is an exemplary sleep prediction graph based on a random forest algorithm according to an embodiment of the present disclosure;
fig. 5 is a general flowchart framework diagram of an exemplary sleep quality assessment method provided by an embodiment of the present application;
fig. 6 is a flowchart of a sleep quality evaluation method according to an embodiment of the present application;
FIG. 7 is a flowchart of an exemplary training sleep quality assessment model provided by an embodiment of the present application;
FIG. 8 is a diagram illustrating an exemplary Federal learning and edge calculation-based training for evaluating sleep quality of a user according to an embodiment of the present application;
fig. 9 is a first schematic structural diagram of a client according to an embodiment of the present disclosure;
fig. 10 is a schematic diagram illustrating a composition structure of a client according to an embodiment of the present application
Fig. 11 is a first schematic structural diagram of a server according to an embodiment of the present disclosure;
fig. 12 is a schematic structural diagram of a server according to an embodiment of the present application.
Detailed Description
The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
The current medical health related application program generally collects and stores a large amount of user sleep monitoring data in a remote cloud center, and performs user health monitoring by using a Machine Learning (ML) model trained by the user sleep monitoring data of the remote cloud center. As shown in fig. 1, each end user may send its own user sleep data (user sleep monitoring data) to a remote cloud center (data processing center), and the data processing center may train a machine learning model in a centralized manner by using the user sleep data corresponding to each end user. Because the sleep monitoring data of the user relates to the privacy of the user, the collection and storage of a large amount of user sleep monitoring data in the remote cloud center brings a great privacy disclosure risk, and the safety of the user sleep quality evaluation is reduced.
Example one
An embodiment of the present application provides a sleep quality assessment method, which is applied to a client, and fig. 2 is a flowchart of a sleep quality assessment method provided in the embodiment of the present application, and as shown in fig. 2, the sleep quality assessment method may include:
s101, sending a model obtaining instruction to the server to obtain the first model from the server based on the model obtaining instruction.
The sleep quality assessment method provided by the embodiment of the application is suitable for a scene that a first model is trained at a client side, and the sleep quality assessment model is used for performing sleep quality assessment on the corresponding to-be-tested environment parameters to be tested.
In the present embodiment, the client may be implemented in various forms. For example, the client described in the present application may include devices such as a mobile phone, a camera, a tablet computer, a notebook computer, a palmtop computer, a Personal Digital Assistant (PDA), a Portable Media Player (PMP), a navigation device, a wearable device, a smart band, a pedometer, a smart gateway, a router, and the like, and devices such as a Digital TV, a desktop computer, and the like.
In the embodiment of the application, the number of the clients may be one, two or more; the specific number of the clients may be determined according to an actual request situation, which is not limited in the embodiment of the present application.
In the embodiment of the application, the client can send the model obtaining instruction to the server at regular time so as to obtain the first model from the server based on the model obtaining instruction; the client side can also send a model obtaining instruction to the server under the condition of obtaining the sample environment parameters and the sample sleep labels corresponding to the sample objects, so as to obtain the first model from the server based on the model obtaining instruction; the client side can also send a model obtaining instruction to the server under other conditions so as to obtain the first model from the server based on the model obtaining instruction; the specific details can be determined according to actual situations, and the embodiment of the present application does not limit the details.
In the embodiment of the application, if the client sends the model acquisition instruction to the server at regular time, the timing time can be 8 points in the morning every day; the timing time can also be 0 pm per day; the timing time may also be other points in time; the specific timing time point may be determined according to actual conditions, which is not limited in the embodiment of the present application.
It should be noted that the first model may be a model with initialized model parameters; the first model can also be a model which is trained by using training environment parameters and training sleep labels corresponding to the training sample object, and the model loss is less than a preset model loss threshold value; the specific first model may be determined according to actual conditions, which is not limited in the embodiment of the present application.
It should be further noted that, if the first model is a model which has been trained by using training environment parameters and training sleep labels corresponding to training sample objects and has a model loss smaller than a preset model loss threshold, the client trains the first model by using sample environment parameters and sample sleep labels corresponding to the sample objects to obtain a sleep quality assessment model, the first model is adjusted by using sample environment parameters and sample sleep labels corresponding to the sample objects at the client, and the adjusted sleep quality assessment model is a model having an ability of analyzing environment parameters of the to-be-tested object corresponding to the client during sleep, so that accuracy of the sleep quality assessment model in assessing a sleep quality assessment result corresponding to the to-be-tested environment parameters at the client is improved.
In the embodiment of the application, the preset model loss threshold may be a threshold configured in the client; the preset model loss threshold value can also be a threshold value in the input value client; the preset model loss threshold value can also be a threshold value transmitted to the client by other devices; the specific manner in which the client acquires the preset model loss threshold may be determined according to actual conditions, which is not limited in the embodiment of the present application.
S102, training a first model by using sample environment parameters and sample sleep labels corresponding to sample objects to obtain a sleep quality evaluation model; the sample environment parameter is the environment parameter of the environment where the sample object is located; the sample sleep label is a sleep label corresponding to the sample environmental parameters.
In the embodiment of the application, after the client sends the model obtaining instruction to the server to obtain the first model from the server based on the model obtaining instruction, the client can train the first model by using the sample environment parameters and the sample sleep labels corresponding to the sample objects to obtain the sleep quality evaluation model.
It should be noted that the sample environmental parameter is an environmental parameter of an environment in which the sample object is located; the sample sleep label is a sleep label corresponding to the sample environmental parameters.
It should be further noted that the sample environment parameters may include: the light intensity of the environment where the sample object is located, the sound amplitude of the environment where the sample object is located, and the motion acceleration change information of the sample object; the environment parameters may further include screen occlusion parameters of the client, gravity change information of the sample object, a client screen on-off state, usage state parameters of the client application, and the like. Specific sample environment parameters can be determined according to actual conditions, and the embodiment of the application does not limit the parameters. An exemplary sample environmental parameter is shown in table 1:
TABLE 1 sample environmental parameter Table
In the embodiment of the application, the client may be a mobile phone, and may acquire sensor data (i.e., sample environmental parameters) acquired by a mobile phone sensor by using an application program, and record and store the sensor data in a log file. The number of sample objects may be 34, and the sample environmental parameter may be an environmental parameter of the 34 sample objects over a period of time (45 days). For each of the 34 sample objects, each client may divide the environment parameters of each sample object in a proportion of 3:7 into test sample environment parameters and training sample environment parameters.
It should be noted that the mobile phone sensor acquisition may acquire the environmental parameters once at time intervals of 10 minutes.
In the embodiment of the application, the client regularly displays the instruction for collecting the sample sleep label every day, so that the sample object inputs a daily sleep diary (sample sleep label) which comprises the sleep-in time, the sleep latency (the time required for falling asleep), the sleep-wake time, the score of the sample object on the sleep quality of the previous night (the score level can be 5 levels including excellent 5, good 4, general 3, poor 2 and poor 1), the sleep disorder (such as the inability to fall asleep within 30 minutes, awakening in the middle of the night, and the like), and the sleep environment, wherein the sleep environment is the position of the mobile phone (on the bed, near the bed, in the bedroom and not in the bedroom) when sleeping. And taking a time window between the wake-up time and the sleep-in time of the sample object as a sleep time, marking the time window as sleep, and marking the parameters of other periods as non-sleep.
Note that the score of sleep quality includes 5 levels: the 5 th grade is excellent, the 4 th grade is good, the 3 rd grade is general, the 2 nd grade is poor, and the 1 st grade is very poor.
It should be noted that the sleep disorder may include a case where the user cannot fall asleep within 30 minutes, or awakening in the middle of the night.
The sample environment parameter is a parameter including time information.
In an embodiment of the present application, the sample object may be a sample user. The sample sleep label comprises parameters such as sleep-in time, sleep-out time, sleep duration, time required for falling asleep, user score for sleep quality, and sleep disorder, which are input by a sample user.
In this embodiment of the present application, a process in which a client trains a first model by using sample environment parameters and sample sleep labels corresponding to sample objects to obtain a sleep quality assessment model includes: the client trains a first model by using the sample environment parameters and the sample sleep label to obtain an initial training model; the client side obtains initial model parameters of an initial training model; sending the initial model parameters to a server; the server adjusts the first model at the server according to the initial model parameters to obtain an adjusted first model; the method comprises the steps that a client side obtains model adjustment parameters corresponding to an adjusted first model from an instruction when receiving the instruction of convergence of the adjusted first model transmitted by a server; and the client replaces the initial model parameters in the initial training model by the model adjustment parameters to obtain an initial model, and the initial model is used as a sleep quality evaluation model.
In the embodiment of the application, under the condition that the number of the clients is at least one, at least one client sends at least one initial model parameter to the server; the process of the server adjusting the first model at the server according to the initial model parameter to obtain the adjusted first model may be: the server carries out weighting processing on at least one initial model parameter to obtain an initial model weighting parameter; and then the server replaces the model parameters in the first model by the initial model weighting parameters to obtain the adjusted first model.
In this embodiment of the application, the adjusted first model converges, and a model loss corresponding to the adjusted first model may be smaller than a preset model loss threshold.
In the embodiment of the application, after the client sends the initial model parameters to the server, the client acquires the model adjustment parameters from the instruction under the condition that the client receives the instruction that the adjusted first model transmitted by the server is not converged; and the client continuously trains the initial training model by using the model adjusting parameters, the sample environment parameters and the sample sleep label until a sleep quality evaluation model is obtained.
In this embodiment of the application, the adjusted first model does not converge, and a model loss corresponding to the adjusted first model may be greater than or equal to a preset model loss threshold.
In the embodiment of the application, the process that the client continues to train the initial training model by using the model adjustment parameter, the sample environment parameter and the sample sleep label until the sleep quality evaluation model is obtained includes: the client replaces initial model parameters in the initial training model by using the model adjustment parameters to obtain an initial model; the client trains an initial model by using the sample environment parameters and the sample sleep label to obtain a training adjustment model; and the client sends the obtained training model parameters of the training adjustment model to the server until the convergence model parameters of the convergence model are obtained from the convergence instruction under the condition that the convergence instruction of the convergence model, which is transmitted by the server and is obtained according to the training model parameters, is received, and the training model parameters in the training adjustment model are replaced according to the convergence model parameters to obtain the sleep quality evaluation model.
In this embodiment of the application, the process of sending the obtained training model parameters of the training adjustment model to the server until obtaining the convergence model parameters of the convergence model from the convergence instruction under the condition of receiving the convergence instruction of the convergence model obtained according to the training model parameters transmitted by the server, and replacing the training model parameters in the training adjustment model according to the convergence model parameters to obtain the sleep quality assessment model may be: the client sends the training model parameters to the server; the server adjusts the adjusted first model at the server according to the training model parameters to obtain a sleep quality evaluation model for continuous training; under the condition that the client receives an instruction that the sleep quality evaluation model for continuous training is a convergence model (namely the sleep quality evaluation model for continuous training is converged), the client acquires a first parameter corresponding to the sleep quality evaluation model for continuous training from the instruction; and the client replaces the training model parameters in the training adjustment model by the first parameters, so that the sleep quality evaluation model is obtained. Under the condition that the client receives an instruction that the sleep quality evaluation model for continuous training is not a convergence model (namely the sleep quality evaluation model for continuous training is not converged), the client acquires a first parameter corresponding to the sleep quality evaluation model for continuous training from the instruction; and continuously training the training adjustment model by using the first parameter, the sample environment parameter and the sample sleep label until a sleep quality evaluation model is obtained.
In this embodiment of the present application, a process in which a client trains a first model by using sample environment parameters and sample sleep tags to obtain an initial training model includes: the client eliminates the first environment parameters without the first sample sleep label from the sample environment parameters to obtain the eliminated sample environment parameters; the client removes a second sleep label without a second sample environment parameter from the sample sleep labels to obtain the removed sample sleep label; and the client trains the first model by using the eliminated sample environment parameters and the eliminated sample sleep labels to obtain an initial training model.
In this embodiment of the application, the first environment parameter is a parameter that the sample object does not input the first sample sleep tag, but acquires the sample environment parameter in the corresponding time period at the client. Illustratively, in the case that the sample object does not input the sleep diary (first sample sleep tag) of the day, i.e., does not have tag information, the corresponding first environmental parameter in the day is rejected.
In this embodiment of the application, the second sleep tag is a tag that the second sample environment parameter corresponding to the sample object is not acquired, but the sample object inputs the sample sleep tag in the corresponding time period. Illustratively, in the absence of sensor data (second sample environmental parameter), the corresponding second sleep tag is rejected during the day.
It can be understood that the client end gets the rejected sample environment parameters by rejecting the first environment parameters without the first sample sleep tag from the sample environment parameters; removing second sleep tags without second sample environment parameters from the sample sleep tags to obtain removed sample sleep tags; the removed sample environmental parameters and the removed sample sleep labels are sample data in one-to-one correspondence, and the removed sample environmental parameters and the removed sample sleep labels in one-to-one correspondence are used for training the first model, so that noise in the initial training model is reduced, and the accuracy of the initial training model is improved.
In this embodiment of the application, the process of training the first model by the client using the removed sample environmental parameters and the removed sample sleep label to obtain the initial training model includes: the client side obtains sample environment characteristics from the eliminated sample environment parameters by using a decision tree method; and the client trains the first model according to the sample environment characteristics, the eliminated sample environment parameters and the eliminated sample sleep label to obtain an initial training model.
For example, as shown in fig. 3, the data acquisition module of the client acquires a variety of sample environment parameters corresponding to a sample object from the terminal (the sample environment parameters include sound amplitude (sound amplitude) of an environment where the sample object is located, motion acceleration change information (acceleration) of the sample object, light intensity (environment light intensity) of the environment where the sample object is located, a client screen on-off state (screen on-off state), a client screen blocking parameter (screen blocking parameter), and the like), and it is necessary to perform data tagging by using a sleep diary record (sample sleep label); cleaning the sample environment parameters and the sample sleep labels (namely, eliminating the first environment parameters without the first sample sleep labels from the sample environment parameters to obtain the eliminated sample environment parameters, eliminating the second sleep labels without the second sample environment parameters from the sample sleep labels to obtain the eliminated sample sleep labels), then evaluating and selecting various types of cleaned sample environment parameters in a machine learning mode, taking every 10 minutes as a time unit for parameter analysis and parameter feature extraction, namely, collecting various sample environment parameters every 10 minutes, and determining the influence condition of various sample environment parameters on the sleep quality. The method can be beneficial to a C4.5 decision tree, and can be used for classifying various sample environment parameters, dividing the various sample environment parameters into two states of sleep and wakefulness, adopting F1-Measure as an evaluation standard, and screening target sample environment parameters, namely the sample environment parameters, from the various sample environment parameters by using a single feature classification and combined feature classification mode. The final sample environmental parameters include ambient light intensity (mean light intensity), acceleration (standard deviation of acceleration change), and sound amplitude (mean sound amplitude). The client stores the ambient light intensity, acceleration and sound amplitude to a database.
In this embodiment of the application, the process of training the first model by the client using the removed sample environmental parameters and the removed sample sleep label to obtain the initial training model includes: the client divides the rejected sample environment parameters into training sample environment parameters and testing sample environment parameters; acquiring a training sample sleep label corresponding to the training sample environmental parameter and a test sample sleep label corresponding to the test sample environmental parameter from the eliminated sample sleep labels; the client side trains a first model by using the training sample environment parameters and the training sample sleep labels according to a random forest prediction model training mode to obtain a first sleep quality evaluation model; the client inputs the environmental parameters of the test sample into the first sleep quality evaluation model to obtain an output sleep evaluation result; determining the model loss of the first sleep quality evaluation model according to the output sleep evaluation result and the test sample sleep label; under the condition that the model loss is smaller than a preset model loss threshold value, the client side takes the first sleep quality evaluation model as an initial training model; and under the condition that the model loss is greater than or equal to the preset model loss threshold, the client continues to train the first sleep quality assessment model to obtain a first training model, and the first training model is used as an initial training model under the condition that the model loss of the first training model is less than the preset model loss threshold.
In the embodiment of the application, the features of three sensors are selected as the feature set of the sleep data set by means of a C4.5 decision tree. The window data is divided into sleep and awake states according to the data records collected every 10 minute time window. A low pass filter is then used over a series of classification windows to eliminate possible sleep state detection errors such as temporal noise or interruption between very quiet and quiescent states to combine accumulated time slices into the duration of sleep, with the time-to-sleep, sleep duration and wake-up time as inputs to assess the quality of the user's sleep.
In the embodiment of the application, the training method of the random forest prediction model is adopted, and the rejected sample environment parameters are divided into the training sample environment parameters (collected data of the first 30 days) and the testing sample environment parameters (collected data of the last 15 days). The random forest classifier randomly selects and extracts K environmental parameters from the environmental parameters of the training samples to generate K training sets, and the process adopts a repeated mode with back placement to extract data. And then generating K classification trees according to K training sets to form a random forest, and finally determining a final output result by a voting method. The random forest prediction model training mode is shown in fig. 4: dividing the sample environment parameters into training parameters and testing parameters; repeated random sampling (sub-sampling) is put back on the training parameters for K times, and K new training sets (a sub-training set 1, a sub-training set 2 and a sub-training set K …) are generated; according to each of the K new training sets, establishing a decision tree (sub-training) by using a training feature set obtained from training parameters to obtain M (M is less than or equal to K) decision tree models (a base model 1, a base model 2 and a base model … M), and forming a random forest by using the M decision tree models to obtain a first sleep quality evaluation model; and determining the prediction results of the M decision tree models, and determining the output result of the first sleep quality evaluation model by adopting a voting method (classification tree) (determining a comprehensive result by adopting a comprehensive voting method).
S103, under the condition that the to-be-tested environment parameters corresponding to the to-be-tested object are received, the to-be-tested environment parameters are input into the sleep quality evaluation model, and a sleep quality evaluation result corresponding to the to-be-tested object is obtained.
In the embodiment of the application, after the client trains the first model by using the sample environment parameters and the sample sleep labels corresponding to the sample objects to obtain the sleep quality assessment model, the client can input the to-be-tested environment parameters into the sleep quality assessment model under the condition that the client receives the to-be-tested environment parameters corresponding to the to-be-tested object, so as to obtain the sleep quality assessment result corresponding to the to-be-tested object.
In the embodiment of the present application, the sample object and the object to be tested may be the same person; the sample object and the object to be tested may not be the same person; the specific details can be determined according to actual situations, and the embodiment of the present application does not limit the details.
It should be noted that, if the sample object and the object to be tested are the same person, the first model needs to be trained by using the sample environment parameters and the sample sleep labels corresponding to the sample object to obtain a sleep quality assessment model, and then the client can perform sleep quality assessment on the environment parameters to be tested by using the sleep quality assessment model under the condition that the client receives the environment parameters to be tested corresponding to the object to be tested.
For example, an overall flow diagram of an exemplary sleep quality assessment method is shown in fig. 5: the client (terminal equipment/edge node) utilizes the data acquisition module to acquire sample environment parameters, the sample environment parameters comprise sound amplitude of the environment where a sample object is located (sound amplitude acquired by a microphone), motion acceleration change information of the sample object (motion state acquired by an acceleration sensor), light intensity of the environment where the sample object is located (light intensity acquired by a light sensor), and screen shielding parameters of the client (screen shielding acquired by a distance sensor), a sleep diary record (sample sleep label) is collected by utilizing the data acquisition APP, the data processing module acquires the sound amplitude of the environment where the sample object is located, the motion acceleration change information of the sample object, the light intensity of the environment where the sample object is located, the screen shielding parameters of the client, and under the condition of the sample sleep label (raw data loading), the client executes a data cleaning pretreatment process, the set client side eliminates the first environment parameters without the first sample sleep label from the sample environment parameters to obtain the eliminated sample environment parameters; removing second sleep tags without second sample environment parameters from the sample sleep tags to obtain removed sample sleep tags; the client acquires sample environment characteristics (sleep data characteristics) from the eliminated sample environment parameters, and trains the first model by using the eliminated sample environment parameters, the sample environment characteristics and the eliminated sample sleep labels to obtain an initial training model. The client acquires initial model parameters (model parameters) of an initial training model; sending initial model parameters to a server (under the condition that the number of the clients is multiple, the number of the initial model parameters is multiple), and carrying out weighting processing on the multiple initial model parameters by the server to obtain initial model weighting parameters; the server adjusts the first model by using the initial model weighting parameters to obtain an adjusted first model (sleep prediction model), and the server determines convergence information of the adjusted first model; the server acquires the adjusted model adjustment parameters of the first model, and transmits convergence information and the model adjustment parameters (model parameters) to the plurality of clients. The method comprises the steps that a client side obtains model adjustment parameters corresponding to an adjusted first model from an instruction when receiving the instruction of convergence of the adjusted first model transmitted by a server; and replacing initial model parameters in the initial training model by using the model adjustment parameters to obtain an initial model, and taking the initial model as a sleep quality evaluation model. The client side obtains model adjustment parameters from the instruction when receiving the instruction that the adjusted first model is not converged and transmitted by the server; and continuously training the initial training model by using the model adjusting parameters, the sample environment parameters and the sample sleep label until a sleep quality evaluation model is obtained.
The client acquires the first model from the server by using the sample environment parameters corresponding to the sample object and the sample sleep label training to obtain the sleep quality evaluation model, and the sample environment parameters of the sample object do not need to be sent to the server, so that the risk of leakage of the sample environment parameters of the sample object or the to-be-tested environment parameters corresponding to the to-be-tested object does not exist, and the safety during sleep quality evaluation is improved.
Example two
An embodiment of the present application further provides a sleep quality assessment method, which is applied to a server, and fig. 6 is a flowchart of a sleep quality assessment method provided in the embodiment of the present application, and as shown in fig. 6, the sleep quality assessment method may include:
s201, under the condition that a model acquisition instruction transmitted by a client is received, acquiring a first model corresponding to the model acquisition instruction.
The sleep quality assessment method provided by the embodiment of the application is suitable for a scene of transmitting the first model to the client, training the first model by using the client, and assessing the sleep quality of the corresponding to-be-tested environment parameter to be tested by using the sleep quality assessment model.
In the embodiments of the present application, the server may be implemented in various forms. For example, the server described in the present application may include devices such as a mobile phone, a camera, a tablet computer, a notebook computer, a palmtop computer, a Personal Digital Assistant (PDA), a Portable Media Player (PMP), a navigation device, a wearable device, a smart band, a pedometer, and the like, and devices such as a Digital TV, a desktop computer, and the like.
In the embodiment of the application, the number of the clients may be one, two or more; the specific number of the clients may be determined according to an actual request situation, which is not limited in the embodiment of the present application.
S202, sending the first model to the client side, so that the client side can train the first model based on the sample environment parameters and the sample sleep labels at the client side to obtain a sleep quality evaluation model.
In this embodiment of the application, after the server acquires the first model corresponding to the model acquisition instruction, the server may send the first model to the client.
In the embodiment of the application, the number of the clients is at least one; the number of initial model parameters is at least one; after the server sends the first model to the client, the server also receives at least one initial model parameter transmitted by at least one client; the server carries out weighting processing on at least one initial model parameter to obtain an initial model weighting parameter; the server adjusts the first model by using the initial model weighting parameters to obtain an adjusted first model; the server determines convergence information of the adjusted first model; convergence information includes convergence or non-convergence; the server obtains the adjusted model adjustment parameters of the first model and transmits convergence information and the model adjustment parameters to at least one client.
It should be noted that at least one client corresponds to at least one initial model parameter one to one.
In the embodiment of the present application, an exemplary process for training a sleep quality assessment model is shown in fig. 7:
and S71, the client trains the first model by using the sample environment parameters and the sample sleep label to obtain an initial training model.
In the embodiment of the application, the client comprises terminal equipment and an edge node; under the condition that sample environment parameters and sample sleep labels are collected, a first model can be trained on the terminal equipment side according to the selection of a user to obtain an initial training model; or training the first model at the edge node side by the selection of the user to obtain an initial training model.
It should be noted that the edge node is an intelligent gateway.
S72, the client side obtains initial model parameters of the initial training model; and sends the initial model parameters to the server.
S73, the server receives at least one initial model parameter transmitted by at least one client.
It should be noted that the number of the clients is at least one, and the number of the corresponding initial model parameters is at least one.
S74, the server carries out weighting processing on at least one initial model parameter to obtain an initial model weighting parameter.
And S75, the server adjusts the first model by using the initial model weighting parameters to obtain the adjusted first model.
S76, the server determines the convergence information of the adjusted first model.
It should be noted that the convergence information includes convergence or non-convergence.
S77, the server obtains the adjusted model adjustment parameters of the first model and transmits convergence information and the model adjustment parameters to at least one client.
And S78, when receiving the command of the convergence of the adjusted first model transmitted by the server, the client acquires the model adjustment parameters corresponding to the adjusted first model from the command.
And S79, replacing the initial model parameters in the initial training model by the client by using the model adjustment parameters to obtain an initial model, and taking the initial model as a sleep quality evaluation model.
In the embodiment of the application, each user has an intelligent gateway system in the home, the intelligent gateway can be used as a distributed edge node, and if N sample users (sample objects) exist, N sample environment parameters can be collected from the edge nodes of the N sample users. Each of the N sample environment parameters includes privacy and sensitive data, such as light intensity of an environment in which the sample object is located, sound amplitude of the environment in which the sample object is located, motion acceleration change information of the sample object, screen shielding parameters of the client, gravity change information of the sample object, a client screen on-off state, use state parameters of the client application program, sleep data, motion records, heart rate records, and the like. As shown in fig. 8: the method comprises the steps that a terminal or an edge node obtains a first model from a server (a remote data processing center); the method comprises the steps of training a first model (a personal training model) by using sample environment parameters (user sleep data) and sample sleep labels corresponding to sample objects at a terminal side to obtain a sleep quality evaluation model, or training the first model (a local model) by using the sample environment parameters and the sample sleep labels corresponding to the sample objects at each edge node (intelligent gateway) in an edge computing network to obtain the sleep quality evaluation model.
The server sends the first model to the client, so that the client can acquire the first model from the server by using the sample environment parameters corresponding to the sample object and the sample sleep label training to obtain the sleep quality evaluation model, the server does not need to acquire the sample environment parameters of the sample object from the client, the risk that the server leaks the sample environment parameters of the sample object or the to-be-tested environment parameters corresponding to the to-be-tested object does not exist, and the safety in sleep quality evaluation is improved.
EXAMPLE III
Based on the same inventive concept of the embodiments, the embodiments of the present application provide a client 1, which corresponds to a sleep quality assessment method; fig. 9 is a schematic structural diagram of a client according to an embodiment of the present application, where the client 1 may include:
a first sending unit 11, configured to send a model acquisition instruction to a server;
a first obtaining unit 12, configured to obtain a first model from the server based on the model obtaining instruction;
the training unit 13 is configured to train the first model by using the sample environment parameters and the sample sleep labels corresponding to the sample objects to obtain a sleep quality evaluation model; the sample environment parameter is the environment parameter of the environment where the sample object is located; the sample sleep label is a sleep label corresponding to the sample environment parameter;
the input unit 14 is configured to, in a case that a to-be-tested environment parameter corresponding to a to-be-tested object is received, input the to-be-tested environment parameter into the sleep quality assessment model, so as to obtain a sleep quality assessment result corresponding to the to-be-tested object.
In some embodiments of the present application, the client further comprises a replacement unit;
the training unit 13 is configured to train the first model by using the sample environment parameters and the sample sleep label to obtain the initial training model;
the first obtaining unit 12 is configured to obtain initial model parameters of the initial training model; under the condition that an instruction of convergence of the adjusted first model transmitted by the server is received, obtaining a model adjusting parameter corresponding to the adjusted first model from the instruction;
the first sending unit 11 is configured to send the initial model parameters to the server; the server adjusts the first model at the server according to the initial model parameters to obtain an adjusted first model;
and the replacing unit is used for replacing the initial model parameters in the initial training model by using the model adjusting parameters to obtain an initial model, and taking the initial model as the sleep quality evaluation model.
In some embodiments of the present application, the first obtaining unit 12 is configured to, in a case that an instruction transmitted by the server that the adjusted first model does not converge is received, obtain the model adjustment parameter from the instruction;
the training unit 13 is configured to continue training the initial training model by using the model adjustment parameter, the sample environment parameter, and the sample sleep label until the sleep quality assessment model is obtained.
In some embodiments of the present application, the replacing unit is configured to replace an initial model parameter in the initial training model with the model adjustment parameter to obtain an initial model; replacing training model parameters in the training adjustment model according to the convergence model parameters to obtain the sleep quality evaluation model;
the training unit 13 is configured to train the initial model by using the sample environment parameters and the sample sleep label to obtain a training adjustment model;
the first sending unit 11 is configured to send the acquired training model parameters of the training adjustment model to the server;
the first obtaining unit 12 is configured to, until a convergence instruction of the convergence model obtained according to the training model parameters is received, obtain the convergence model parameters of the convergence model from the convergence instruction.
In some embodiments of the present application, the client further comprises a culling unit;
the eliminating unit is used for eliminating the first environment parameters without the first sample sleep label from the sample environment parameters to obtain the eliminated sample environment parameters; removing second sleep tags without the second sample environment parameters from the sample sleep tags to obtain removed sample sleep tags;
the training unit 13 is configured to train the first model by using the eliminated sample environment parameters and the eliminated sample sleep labels, so as to obtain the initial training model.
In some embodiments of the present application, the client further comprises a dividing unit and a first determining unit;
the dividing unit is used for dividing the rejected sample environment parameters into training sample environment parameters and testing sample environment parameters;
the first obtaining unit 12 is configured to obtain, from the eliminated sample sleep labels, a training sample sleep label corresponding to the training sample environment parameter and a test sample sleep label corresponding to the test sample environment parameter;
the training unit 13 is configured to train the first model by using the training sample environment parameters and the training sample sleep labels according to a random forest prediction model training mode to obtain a first sleep quality evaluation model; under the condition that the model loss is greater than or equal to a preset model loss threshold value, continuing to train the first sleep quality assessment model to obtain a first training model, and taking the first training model as the initial training model under the condition that the model loss of the first training model is less than the preset model loss threshold value; taking the first sleep quality evaluation model as the initial training model under the condition that the model loss is less than a preset model loss threshold;
an input unit 14, configured to input the test sample environment parameter into the first sleep quality assessment model, so as to obtain an output sleep assessment result;
the first determining unit is used for determining the model loss of the first sleep quality evaluation model according to the output sleep evaluation result and the test sample sleep label.
In practical applications, the first sending Unit 11, the first obtaining Unit 12, the training Unit 13, and the input Unit 14 may be implemented by a first processor 15 on the client 1, specifically implemented by a CPU (Central Processing Unit), an MPU (micro processor Unit), a DSP (Digital Signal processor), a Field Programmable Gate Array (FPGA), or the like; the above data storage may be implemented by the first memory 16 on the client 1.
An embodiment of the present application further provides a client 1, as shown in fig. 10, where the client 1 includes: a first processor 15, a first memory 16 and a first communication bus 17, the first memory 16 communicating with the first processor 15 through the first communication bus 17, the first memory 16 storing a program executable by the first processor 15, the sleep quality assessment method as described above being performed by the first processor 15 when the program is executed.
In practical applications, the first Memory 16 may be a volatile Memory (volatile Memory), such as a Random-Access Memory (RAM); or a non-volatile Memory (non-volatile Memory), such as a Read-Only Memory (ROM), a flash Memory (flash Memory), a Hard Disk (Hard Disk Drive, HDD) or a Solid-State Drive (SSD); or a combination of the above types of memories and provides instructions and data to the first processor 15.
The present embodiment provides a computer-readable storage medium, on which a computer program is stored, which when executed by the first processor 15 implements the sleep quality assessment method as described above.
The client acquires the first model from the server by using the sample environment parameters corresponding to the sample object and the sample sleep label training to obtain the sleep quality evaluation model, and the sample environment parameters of the sample object do not need to be sent to the server, so that the risk of leakage of the sample environment parameters of the sample object or the to-be-tested environment parameters corresponding to the to-be-tested object does not exist, and the safety during sleep quality evaluation is improved.
Example four
Based on the same inventive concept of the second embodiment, the embodiment of the present application provides a server 2, which corresponds to a sleep quality evaluation method; fig. 11 is a schematic structural diagram of a server according to an embodiment of the present application, where the server 2 may include:
a second obtaining unit 21, configured to, in a case where a model obtaining instruction transmitted by a client is received, obtain a first model corresponding to the model obtaining instruction;
a second sending unit 22, configured to send the first model to the client, so that the client trains the first model based on sample environment parameters and sample sleep tags at the client, and obtains the sleep quality assessment model.
In some embodiments of the present application, the number of clients is at least one; the number of the initial model parameters is at least one; the server also comprises a receiving unit, a processing unit, an adjusting unit and a determining unit;
the receiving unit is used for receiving at least one initial model parameter transmitted by at least one client; the at least one client corresponds to the at least one initial model parameter one by one;
the processing unit is used for weighting the at least one initial model parameter to obtain an initial model weighting parameter;
the adjusting unit is used for adjusting the first model by using the initial model weighting parameter to obtain an adjusted first model;
a determining unit, configured to determine convergence information of the adjusted first model; the convergence information comprises convergence or non-convergence;
a second obtaining unit 21, configured to obtain a model adjustment parameter of the adjusted first model;
a second sending unit 22, configured to transmit the convergence information and the model adjustment parameter to the at least one client.
In practical applications, the second obtaining Unit 21 and the second sending Unit 22 may be implemented by a second processor 23 on the server 2, specifically implemented by a Central Processing Unit (CPU), an MPU (Microprocessor), a Digital Signal Processor (DSP), a Field Programmable Gate Array (FPGA), or the like; the above data storage may be implemented by the second memory 24 on the server 2.
An embodiment of the present application further provides a server 2, as shown in fig. 12, where the server 2 includes: a second processor 23, a second memory 24 and a second communication bus 25, wherein the second memory 24 communicates with the second processor 23 through the second communication bus 25, the second memory 24 stores a program executable by the second processor 23, and when the program is executed, the sleep quality assessment method as described above is executed by the second processor 23.
In practical applications, the second Memory 24 may be a volatile Memory (volatile Memory), such as a Random-Access Memory (RAM); or a non-volatile Memory (non-volatile Memory), such as a Read-Only Memory (ROM), a flash Memory (flash Memory), a Hard Disk (Hard Disk Drive, HDD) or a Solid-State Drive (SSD); or a combination of the above types of memories and provides instructions and data to the second processor 23.
The present embodiment provides a computer-readable storage medium, on which a computer program is stored, which when executed by the second processor 23 implements the sleep quality assessment method as described above.
The server sends the first model to the client, so that the client can acquire the first model from the server by using the sample environment parameters corresponding to the sample object and the sample sleep label training to obtain the sleep quality evaluation model, the server does not need to acquire the sample environment parameters of the sample object from the client, the risk that the server leaks the sample environment parameters of the sample object or the to-be-tested environment parameters corresponding to the to-be-tested object does not exist, and the safety in sleep quality evaluation is improved.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of a hardware embodiment, a software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The above description is only a preferred embodiment of the present application, and is not intended to limit the scope of the present application.
Claims (13)
1. A sleep quality assessment method applied to a client side is characterized by comprising the following steps:
sending a model obtaining instruction to a server to obtain a first model from the server based on the model obtaining instruction;
training the first model by using sample environment parameters and sample sleep labels corresponding to the sample objects to obtain a sleep quality evaluation model; the sample environment parameter is the environment parameter of the environment where the sample object is located; the sample sleep label is a sleep label corresponding to the sample environment parameter;
and under the condition of receiving the to-be-tested environment parameters corresponding to the to-be-tested object, inputting the to-be-tested environment parameters into the sleep quality evaluation model to obtain a sleep quality evaluation result corresponding to the to-be-tested object.
2. The method of claim 1, wherein training the first model using sample environmental parameters and sample sleep labels corresponding to sample objects to obtain a sleep quality assessment model comprises:
training the first model by using the sample environment parameters and the sample sleep label to obtain the initial training model;
acquiring initial model parameters of the initial training model; and sending the initial model parameters to the server; the server adjusts the first model at the server according to the initial model parameters to obtain an adjusted first model;
under the condition that an instruction of convergence of the adjusted first model transmitted by the server is received, obtaining a model adjusting parameter corresponding to the adjusted first model from the instruction;
and replacing initial model parameters in the initial training model by using the model adjustment parameters to obtain an initial model, and taking the initial model as the sleep quality evaluation model.
3. The method of claim 2, wherein after sending the initial model parameters to the server, the method further comprises:
under the condition that an instruction that the adjusted first model is not converged and transmitted by the server is received, acquiring the model adjusting parameters from the instruction;
and continuously training the initial training model by using the model adjusting parameters, the sample environment parameters and the sample sleep label until the sleep quality evaluation model is obtained.
4. The method of claim 3, wherein the continuing training of the initial training model with the model adjustment parameters, the sample environment parameters, and the sample sleep labels until the sleep quality assessment model is obtained comprises:
replacing initial model parameters in the initial training model by the model adjustment parameters to obtain the initial model;
training the initial model by using the sample environment parameters and the sample sleep label to obtain a training adjustment model;
and sending the obtained training model parameters of the training adjustment model to the server until the convergence model parameters of the convergence model are obtained from the convergence instruction under the condition that the convergence instruction of the convergence model, which is transmitted by the server and is obtained according to the training model parameters, is received, and replacing the training model parameters in the training adjustment model according to the convergence model parameters to obtain the sleep quality evaluation model.
5. The method of claim 2, wherein the training the first model using the sample environmental parameters and sample sleep tags, resulting in the initial training model, comprises:
removing the first environmental parameters without the first sample sleep label from the sample environmental parameters to obtain the removed sample environmental parameters;
removing second sleep tags without the second sample environment parameters from the sample sleep tags to obtain removed sample sleep tags;
and training the first model by using the eliminated sample environment parameters and the eliminated sample sleep label to obtain the initial training model.
6. The method of claim 5, wherein the training the first model using the culled sample environment parameters and the culled sample sleep labels to obtain the initial training model comprises:
dividing the rejected sample environment parameters into training sample environment parameters and testing sample environment parameters; acquiring a training sample sleep label corresponding to the training sample environmental parameter and a test sample sleep label corresponding to the test sample environmental parameter from the eliminated sample sleep labels;
training the first model by using the training sample environment parameters and the training sample sleep label according to a random forest prediction model training mode to obtain a first sleep quality evaluation model;
inputting the test sample environmental parameters into the first sleep quality evaluation model to obtain an output sleep evaluation result; determining the model loss of the first sleep quality evaluation model according to the output sleep evaluation result and the test sample sleep label;
taking the first sleep quality evaluation model as the initial training model under the condition that the model loss is less than a preset model loss threshold;
and under the condition that the model loss is greater than or equal to a preset model loss threshold value, continuing to train the first sleep quality assessment model to obtain a first training model, and taking the first training model as the initial training model under the condition that the model loss of the first training model is less than the preset model loss threshold value.
7. A sleep quality assessment method applied to a server, the method comprising:
under the condition of receiving a model acquisition instruction transmitted by a client, acquiring a first model corresponding to the model acquisition instruction;
and sending the first model to the client, so that the client trains the first model based on the sample environment parameters and the sample sleep labels at the client to obtain the sleep quality evaluation model.
8. The method of claim 7, wherein the number of clients is at least one; the number of the initial model parameters is at least one; after the sending the first model to the client, the method further includes:
receiving at least one initial model parameter transmitted by at least one client; the at least one client corresponds to the at least one initial model parameter one by one;
weighting the at least one initial model parameter to obtain an initial model weighting parameter;
adjusting the first model by using the initial model weighting parameter to obtain an adjusted first model;
determining convergence information of the adjusted first model; the convergence information comprises convergence or non-convergence;
and obtaining the model adjusting parameters of the adjusted first model, and transmitting the convergence information and the model adjusting parameters to the at least one client.
9. A client, the client comprising:
the first sending unit is used for sending a model acquisition instruction to the server;
a first obtaining unit, configured to obtain a first model from the server based on the model obtaining instruction;
the training unit is used for training the first model by using the sample environment parameters and the sample sleep labels corresponding to the sample objects to obtain a sleep quality evaluation model; the sample environment parameter is the environment parameter of the environment where the sample object is located; the sample sleep label is a sleep label corresponding to the sample environment parameter;
and the input unit is used for inputting the environmental parameters to be tested into the sleep quality evaluation model under the condition of receiving the environmental parameters to be tested corresponding to the object to be tested, so as to obtain the sleep quality evaluation result corresponding to the object to be tested.
10. A server, characterized in that the server comprises:
the second acquisition unit is used for acquiring a first model corresponding to a model acquisition instruction under the condition of receiving the model acquisition instruction transmitted by the client;
and the second sending unit is used for sending the first model to the client so that the client trains the first model based on the sample environment parameters and the sample sleep labels at the client to obtain the sleep quality evaluation model.
11. A client, the client comprising:
a first memory, a first processor, and a first communication bus, the first memory in communication with the first processor through the first communication bus, the first memory storing a program for sleep quality assessment executable by the first processor, the program for sleep quality assessment, when executed, executing the method of any of claims 1 to 6 through the first processor.
12. A server, characterized in that the server comprises:
a second memory, a second processor, and a second communication bus, the second memory communicating with the second processor through the second communication bus, the second memory storing a program for sleep quality assessment executable by the second processor, the method of any of claims 7 to 8 being performed by the second processor when the program for sleep quality assessment is executed.
13. A storage medium on which a computer program is stored, for application to a client and a server, characterized in that the computer program, when executed by a first processor, implements the method of any one of claims 1 to 6; the computer program, when executed by the second processor, implements the method of any of claims 7 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111602630.2A CN114343574A (en) | 2021-12-24 | 2021-12-24 | Sleep quality evaluation method, client, server and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111602630.2A CN114343574A (en) | 2021-12-24 | 2021-12-24 | Sleep quality evaluation method, client, server and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114343574A true CN114343574A (en) | 2022-04-15 |
Family
ID=81101104
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111602630.2A Pending CN114343574A (en) | 2021-12-24 | 2021-12-24 | Sleep quality evaluation method, client, server and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114343574A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114947458A (en) * | 2022-06-27 | 2022-08-30 | 慕思健康睡眠股份有限公司 | Hotel room recommendation method, device, equipment and medium based on sleep mattress |
| CN115496105A (en) * | 2022-09-28 | 2022-12-20 | 广东省新黄埔中医药联合创新研究院 | Sleep prediction model training method, sleep condition prediction method and related device |
| CN117077812A (en) * | 2023-09-13 | 2023-11-17 | 荣耀终端有限公司 | Network training method, sleep state evaluation method and related equipment |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107007263A (en) * | 2017-04-25 | 2017-08-04 | 中国科学院计算技术研究所 | The sleep quality assessment method and system of a kind of generalization |
| KR102253946B1 (en) * | 2020-08-04 | 2021-05-20 | 주식회사 에이슬립 | Computer device for prediction sleep state based on data measured in user's sleep environment |
| CN113111952A (en) * | 2021-04-20 | 2021-07-13 | 北京心康医学科技有限公司 | Sleep training method and system based on artificial intelligence |
-
2021
- 2021-12-24 CN CN202111602630.2A patent/CN114343574A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107007263A (en) * | 2017-04-25 | 2017-08-04 | 中国科学院计算技术研究所 | The sleep quality assessment method and system of a kind of generalization |
| KR102253946B1 (en) * | 2020-08-04 | 2021-05-20 | 주식회사 에이슬립 | Computer device for prediction sleep state based on data measured in user's sleep environment |
| CN113111952A (en) * | 2021-04-20 | 2021-07-13 | 北京心康医学科技有限公司 | Sleep training method and system based on artificial intelligence |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114947458A (en) * | 2022-06-27 | 2022-08-30 | 慕思健康睡眠股份有限公司 | Hotel room recommendation method, device, equipment and medium based on sleep mattress |
| CN114947458B (en) * | 2022-06-27 | 2023-09-05 | 慕思健康睡眠股份有限公司 | Hotel room recommending method, device, equipment and medium based on sleeping mattress |
| CN115496105A (en) * | 2022-09-28 | 2022-12-20 | 广东省新黄埔中医药联合创新研究院 | Sleep prediction model training method, sleep condition prediction method and related device |
| CN115496105B (en) * | 2022-09-28 | 2023-10-24 | 广东省新黄埔中医药联合创新研究院 | Sleep prediction model training method, sleep condition prediction method and related devices |
| CN117077812A (en) * | 2023-09-13 | 2023-11-17 | 荣耀终端有限公司 | Network training method, sleep state evaluation method and related equipment |
| CN117077812B (en) * | 2023-09-13 | 2024-03-08 | 荣耀终端有限公司 | Network training method, sleep state evaluation method and related equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114343574A (en) | Sleep quality evaluation method, client, server and storage medium | |
| CN107773214A (en) | A kind of method, computer-readable medium and the system of optimal wake-up strategy | |
| CN104143083B (en) | A kind of face identification system of Kernel-based methods management | |
| CN104142583A (en) | Intelligent glasses with blinking detection function and implementation method thereof | |
| CN109363659B (en) | Heart rate monitoring method and device based on deep learning and storage medium | |
| CN109276255A (en) | A kind of limb tremor detection method and device | |
| CN104490401A (en) | Method for intelligent terminal for automatically recognizing human behaviors based on acceleration data | |
| CN104539726A (en) | System and method used for concerning behavior rules of the old by children | |
| CN103845038A (en) | Physical sign signal acquiring method and physical sign signal acquiring equipment | |
| US20230255487A1 (en) | System and method for biometric-based spatial categorization | |
| CN113951879B (en) | Blood glucose prediction method and device and system for monitoring blood glucose level | |
| CN111790038A (en) | Music sleep assisting method and device, computer equipment and storage medium | |
| CN108937866A (en) | Dormant monitoring method and device | |
| CN107019501B (en) | Remote tumble detection method and system based on genetic algorithm and probabilistic neural network | |
| CN106821349A (en) | For the event generation method and device of wearable custodial care facility | |
| CN109300546A (en) | A kind of individual sub-health state appraisal procedure based on big data and artificial intelligence | |
| CN112183638A (en) | Hard rock tensile-shear fracture identification method and device based on voiceprint depth characteristics | |
| CN117322852A (en) | Intelligent clothing health monitoring system and method based on deep learning | |
| Yang et al. | Loneliness forecasting using multi-modal wearable and mobile sensing in everyday settings | |
| CN115299943A (en) | Emotion monitoring method and device | |
| CN106940720A (en) | Multi-source information processing method and system based on healthy Internet of Things | |
| CN108461125A (en) | A kind of memory training device for the elderly | |
| CN114048849A (en) | Unsupervised representation learning method and device supporting time sequence social network diagram | |
| Barrett et al. | Mobile sensing: Leveraging machine learning for efficient human behavior modeling | |
| Bhuvaneswari et al. | Identifying event bursts using log-normal distribution of tweet arrival rate in twitter stream |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |