CN115547457B - Recipe intelligent recommendation method, device, equipment and medium based on physical examination data - Google Patents
Recipe intelligent recommendation method, device, equipment and medium based on physical examination data Download PDFInfo
- Publication number
- CN115547457B CN115547457B CN202211260581.3A CN202211260581A CN115547457B CN 115547457 B CN115547457 B CN 115547457B CN 202211260581 A CN202211260581 A CN 202211260581A CN 115547457 B CN115547457 B CN 115547457B
- Authority
- CN
- China
- Prior art keywords
- recipe
- representing
- physical examination
- vector
- data
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 43
- 239000013598 vector Substances 0.000 claims abstract description 288
- 238000012545 processing Methods 0.000 claims abstract description 18
- 238000006243 chemical reaction Methods 0.000 claims abstract description 11
- 230000007246 mechanism Effects 0.000 claims description 19
- 230000006870 function Effects 0.000 claims description 14
- 238000004590 computer program Methods 0.000 claims description 11
- 238000004422 calculation algorithm Methods 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 4
- 238000010276 construction Methods 0.000 abstract 1
- 238000013136 deep learning model Methods 0.000 description 10
- 238000004891 communication Methods 0.000 description 8
- 238000004458 analytical method Methods 0.000 description 7
- 230000008569 process Effects 0.000 description 7
- 238000007726 management method Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 230000001788 irregular Effects 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 239000008280 blood Substances 0.000 description 3
- 210000004369 blood Anatomy 0.000 description 3
- 238000007405 data analysis Methods 0.000 description 3
- 238000013135 deep learning Methods 0.000 description 3
- 230000003993 interaction Effects 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 230000036772 blood pressure Effects 0.000 description 2
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 2
- 239000003814 drug Substances 0.000 description 2
- 239000004973 liquid crystal related substance Substances 0.000 description 2
- 210000000214 mouth Anatomy 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000002093 peripheral effect Effects 0.000 description 2
- 210000002784 stomach Anatomy 0.000 description 2
- 238000001356 surgical procedure Methods 0.000 description 2
- WQZGKKKJIJFFOK-GASJEMHNSA-N Glucose Natural products OC[C@H]1OC(O)[C@H](O)[C@@H](O)[C@@H]1O WQZGKKKJIJFFOK-GASJEMHNSA-N 0.000 description 1
- 235000007688 Lycopersicon esculentum Nutrition 0.000 description 1
- 240000003768 Solanum lycopersicum Species 0.000 description 1
- 208000005718 Stomach Neoplasms Diseases 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004159 blood analysis Methods 0.000 description 1
- 210000001124 body fluid Anatomy 0.000 description 1
- 239000010839 body fluid Substances 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000035487 diastolic blood pressure Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000000802 evaporation-induced self-assembly Methods 0.000 description 1
- 235000013305 food Nutrition 0.000 description 1
- 206010017758 gastric cancer Diseases 0.000 description 1
- 239000008103 glucose Substances 0.000 description 1
- 230000001900 immune effect Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 238000007637 random forest analysis Methods 0.000 description 1
- 230000028327 secretion Effects 0.000 description 1
- 201000011549 stomach cancer Diseases 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
- 230000035488 systolic blood pressure Effects 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 210000002700 urine Anatomy 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H20/00—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance
- G16H20/60—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance relating to nutrition control, e.g. diets
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/70—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Public Health (AREA)
- Primary Health Care (AREA)
- General Health & Medical Sciences (AREA)
- Epidemiology (AREA)
- Data Mining & Analysis (AREA)
- Nutrition Science (AREA)
- Biomedical Technology (AREA)
- Databases & Information Systems (AREA)
- Pathology (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention relates to the field of intelligent decision making, and discloses a recipe intelligent recommendation method, device, electronic equipment and storage medium based on physical examination data, wherein the method comprises the following steps: obtaining physical examination data and recipe data, identifying physical examination values and physical examination categories in the physical examination data, constructing numerical vectors and category vectors, performing dense conversion on the numerical vectors and the category vectors, and determining recipe vectors corresponding to the recipe data; calculating a correlation coefficient between the dense vector and the recipe vector, calculating a cross feature between the dense vector and the recipe vector according to the correlation coefficient, and determining a feature score; performing deep cross processing on the feature cascade by utilizing the feature score and the feature cascade of the dense vector construction physical examination data to obtain a deep cross cascade, and calculating recipe recommendation probability corresponding to the deep cross cascade; and determining a recipe recommendation result of the physical examination data in the recipe data according to the recipe recommendation probability. The method and the device can improve the intelligent recommending depth of the recipes based on physical examination data.
Description
Technical Field
The invention relates to the field of intelligent decision making, in particular to a recipe intelligent recommendation method and device based on physical examination data, electronic equipment and a storage medium.
Background
The intelligent recipe recommendation based on physical examination data refers to a process of taking physical examination data as input of a deep learning model and outputting recommendation probabilities of different types of recipes by using the model.
At present, the existing recipe recommendation algorithm comprises a random forest support vector machine, a deep neural network and other recommendation algorithms, but the methods are all based on numerical domain data such as physical examination indexes and the like to construct a model, so that the influence of text domain data such as physical examination body types (such as senile type, middle-aged type and the like), physical examination item types (such as gastric cancer physical examination, conventional physical examination and the like) and the like is greatly ignored, the text domain data are basic properties of physical examination, and can reflect the direction of a recipe required to be recommended by a physical examination person, such as physical examination on stomach, and represent that the physical examination person needs to be recommended to be related to stomach, in addition, the recipe requirements of different physical types are different, for example, the recipe requirements of a child physical examination person are mild recipes; and there is insufficient attention to the inherent correlation between physical examination data. Therefore, the intelligent recipe recommendation depth based on physical examination data is insufficient.
Disclosure of Invention
In order to solve the problems, the invention provides a recipe intelligent recommendation method, device, electronic equipment and storage medium based on physical examination data, which can improve the recipe intelligent recommendation depth based on physical examination data.
In a first aspect, the present invention provides an intelligent recipe recommendation method based on physical examination data, including:
acquiring physical examination data and recipe data, identifying physical examination values and physical examination categories in the physical examination data, constructing numerical vectors and category vectors corresponding to the physical examination values and the physical examination categories, performing dense conversion on the numerical vectors and the category vectors to obtain dense vectors, and determining recipe vectors corresponding to the recipe data;
calculating a correlation coefficient between the dense vector and the recipe vector, calculating a cross feature between the dense vector and the recipe vector according to the correlation coefficient, and determining a feature score corresponding to the cross feature;
constructing a feature cascade of the physical examination data by using the feature score and the dense vector, performing deep cross processing on the feature cascade to obtain a deep cross cascade, and calculating a recipe recommendation probability corresponding to the deep cross cascade;
And determining a recipe recommendation result corresponding to the physical examination data in the recipe data according to the recipe recommendation probability.
In a possible implementation manner of the first aspect, the constructing a numeric vector and a category vector of the physical examination numeric value and the physical examination category includes:
inquiring the numerical name of the physical examination numerical value;
constructing a name vector and a category vector corresponding to the numerical name and the physical examination category by using a single-hot encoding algorithm;
calculating the numerical code of the physical examination value by using the following formula:
wherein σ represents the numerical code of the physical examination value, and x represents the physical examination value;
and vector splicing is carried out on the numerical code and the name vector, so that a numerical vector corresponding to the physical examination numerical value is obtained.
In a possible implementation manner of the first aspect, the calculating the association coefficient between the dense vector and the recipe vector includes:
calculating an association weight between the dense vector and the recipe vector using the formula:
wherein a is ij Representing the associated weights, dense representing the fully connected layer,
represents the dense vector, pe represents the physical examination data, T represents a transpose symbol, I re Representing the recipe vector, re representing the recipe data, i representing the serial number of the physical examination data, j representing the serial number of the recipe data;
determining an association coefficient between the dense vector and the recipe vector according to the association weight using the following formula:
wherein Weight is attention Representing the correlation coefficient between the dense vector and the recipe vector, the attention represents the attention mechanism layer, a 11 、a 1J 、a I1 、a IJ The association weight I, J indicates the number of serial numbers of the physical examination data and the number of serial numbers of the recipe data.
In a possible implementation manner of the first aspect, the calculating a cross feature between the dense vector and the recipe vector according to the association coefficient includes:
calculating a dense cross feature of the dense vector from the correlation coefficient using the formula:
wherein βi represents a dense cross feature of the dense vector, J represents the number of sequence numbers of the recipe data, pe represents the physical examination data, i represents the sequence number of the physical examination data, J represents the sequence number of the recipe data,
representing an ith dense vector corresponding to the physical examination data;
according to the correlation coefficient, the recipe intersection characteristic of the recipe vector is calculated by using the following formula:
Wherein alpha is j A recipe cross feature representing the recipe vector, I representing the number of serial numbers of the physical examination data, re representing the recipe data, I representing the serial number of the physical examination data, j representing the serial number of the recipe data,
a j-th recipe vector representing the recipe data;
calculating the intersection feature between the dense vector and the recipe vector from the dense intersection feature and the recipe intersection feature using the following formula:
wherein X is attention Representing the intersection features between the dense vector and the recipe vector, alpha representing the recipe intersection features of the recipe vector, beta representing the dense intersection features of the dense vector,
represents the dense vector, pe represents the physical examination data, T represents a transpose symbol, I re Representing the recipe vector, re representing the recipe data.
In a possible implementation manner of the first aspect, the constructing a feature cascade of the physical examination data using the feature score and the dense vector includes:
the feature cascade of the physical examination data is constructed using the following formula:
wherein x is 0 Representing the cascade of features, X' attention Representing the feature score corresponding to the cross feature, U representing the dense vector, T representing the transposed symbol, emmbed, 1 representing X' attention And the row and column of U, k represents X' attention And the number of columns of U.
In a possible implementation manner of the first aspect, the performing a deep cross process on the feature cascade to obtain a deep cross cascade includes:
performing multi-layer cross processing on the characteristic cascade by using the following formula to obtain a multi-layer cross cascade:
wherein X' attention Representing the multi-layer cross-cascade, x 0 Representing the cascade of features, W attention Network weights representing the attention mechanism layer, b attention Representing parameter deviation of the attention mechanism layer, X attention Representing the dense vectorThe cross feature between the recipe vectors, T representing the transposed symbol;
the feature depth combination of the feature cascade is constructed using the following formula:
h′ attention ＝f(w attention h attention +b attention )
wherein h' attention A feature depth combination representing the feature cascade, W attention Network weights representing the attention mechanism layer, b attention Indicating parameter deviation of the attention mechanism layer, h attention Representing a fully connected layer of the depth network layer, f representing a function of the depth network layer;
and determining the depth cross cascade of the feature cascade according to the multi-layer cross cascade and the feature depth combination.
In a possible implementation manner of the first aspect, the determining, according to the recipe recommendation probability, a recipe recommendation result corresponding to the physical examination data in the recipe data includes:
Extracting a target recommendation probability which is larger than a preset probability from the recipe recommendation probability;
inquiring target recipe data corresponding to the target recommendation probability in the recipe data;
and carrying out recipe data combination on the target recipe data to obtain a recipe recommendation result corresponding to the physical examination data.
In a second aspect, the present invention provides an intelligent recipe recommendation device based on physical examination data, the device comprising:
the recipe vector determining module is used for acquiring physical examination data and recipe data, identifying physical examination values and physical examination categories in the physical examination data, constructing numerical vectors and category vectors corresponding to the physical examination values and the physical examination categories, performing dense conversion on the numerical vectors and the category vectors to obtain dense vectors, and determining recipe vectors corresponding to the recipe data;
the characteristic score determining module is used for calculating a correlation coefficient between the dense vector and the recipe vector, calculating a cross characteristic between the dense vector and the recipe vector according to the correlation coefficient, and determining a characteristic score corresponding to the cross characteristic;
the recommendation probability calculation module is used for constructing a feature cascade of the physical examination data by utilizing the feature score and the dense vector, carrying out deep cross processing on the feature cascade to obtain a deep cross cascade, and calculating a recipe recommendation probability corresponding to the deep cross cascade;
And the recommended result determining module is used for determining a recipe recommended result corresponding to the physical examination data in the recipe data according to the recipe recommended probability.
In a third aspect, the present invention provides an electronic device comprising:
at least one processor; and a memory communicatively coupled to the at least one processor;
wherein the memory stores a computer program executable by the at least one processor to enable the at least one processor to perform the intelligent recipe recommendation method based on physical examination data as described in any one of the first aspects above.
In a fourth aspect, the present invention provides a computer readable storage medium storing a computer program which when executed by a processor implements the intelligent recipe recommendation method based on physical examination data as described in any one of the first aspects.
Compared with the prior art, the technical principle and beneficial effect of this scheme lie in:
the embodiment of the invention firstly recommends proper recipes according to physical examination results by acquiring physical examination data and recipe data, further, the embodiment of the invention reduces the turbulence of data during subsequent data analysis by identifying physical examination values and physical examination categories in the physical examination data for preliminary classification, further, the embodiment of the invention facilitates the subsequent analysis of data by using a deep learning model by constructing numerical vectors and category vectors corresponding to the physical examination values and the physical examination categories, facilitates the subsequent analysis of data by using a deep learning model, further, the embodiment of the invention facilitates the computation of the correlation between the numerical vectors and the recipe data by using a deep learning model, further, the embodiment of the invention adapts to the structural characteristics of deep learning by using sparse vectors to be converted into dense vectors, further, the embodiment of the invention facilitates the correlation between the recipe data by determining the recipe vectors corresponding to the recipe data, facilitates the subsequent analysis by using a deep learning model, further, facilitates the correlation between the recipe vectors and the recipe data by using the deep learning model, further, the embodiment of the invention constructs a feature cascade of the physical examination data by using the feature score and the dense vector to be used for carrying out cross combination on the dense vector and the cross feature by determining the feature score corresponding to the cross feature to be used for determining the attention score corresponding to the cross feature, namely the matching score of the physical examination data and the corresponding recipe data, further, the embodiment of the invention improves the accuracy of matching the physical examination data and the recipe data by carrying out deep cross processing on the feature cascade to be used for improving the interaction degree between the dense vector and the recipe vector, further, the embodiment of the invention improves the accuracy of matching the physical examination data and the recipe data by calculating the recipe recommendation probability corresponding to the deep cross cascade to be used for marking the recipe data corresponding to the physical examination data by using the probability, and further, the embodiment of the invention determines a recipe recommendation result corresponding to the physical examination data in the recipe data by using the recipe recommendation probability according to the recipe recommendation probability to be used for determining the recipe recommendation corresponding to the physical examination data according to the output result of the model, and further, the embodiment of the invention reduces the intelligent manpower problem by using the model. Therefore, the recipe intelligent recommendation method, device, electronic equipment and storage medium based on the physical examination data can improve the recipe intelligent recommendation depth based on the physical examination data.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and together with the description, serve to explain the principles of the invention.
In order to more clearly illustrate the embodiments of the invention or the technical solutions of the prior art, the drawings which are used in the description of the embodiments or the prior art will be briefly described, and it will be obvious to a person skilled in the art that other drawings can be obtained from these drawings without inventive effort.
Fig. 1 is a schematic flow chart of a recipe intelligent recommendation method based on physical examination data according to an embodiment of the present invention;
FIG. 2 is a flowchart illustrating one of the steps of the intelligent recipe recommendation method based on physical examination data according to an embodiment of the present invention;
FIG. 3 is a flowchart illustrating another step of the intelligent recipe recommendation method based on physical examination data according to an embodiment of the present invention;
fig. 4 is a schematic block diagram of a recipe intelligent recommendation device based on physical examination data according to an embodiment of the present invention;
fig. 5 is a schematic diagram of an internal structure of an electronic device for implementing a recipe intelligent recommendation method based on physical examination data according to an embodiment of the present invention.
Detailed Description
It should be understood that the detailed description is presented by way of example only and is not intended to limit the invention.
The embodiment of the invention provides a recipe intelligent recommendation method based on physical examination data, wherein an execution subject of the recipe intelligent recommendation method based on physical examination data comprises, but is not limited to, at least one of a server, a terminal and the like which can be configured to execute the method provided by the embodiment of the invention. In other words, the intelligent recipe recommendation method based on physical examination data can be executed by software or hardware installed in a terminal device or a server device, and the software can be a blockchain platform. The service end includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, and the like. The server may be an independent server, or may be a cloud server that provides cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (Content Delivery Network, CDN), and basic cloud computing services such as big data and artificial intelligence platforms.
Referring to fig. 1, a flow chart of an intelligent recipe recommendation method based on physical examination data according to an embodiment of the invention is shown. The recipe intelligent recommendation method based on physical examination data depicted in fig. 1 comprises the following steps:
s1, acquiring physical examination data and recipe data, identifying physical examination values and physical examination categories in the physical examination data, constructing numerical vectors and category vectors corresponding to the physical examination values and the physical examination categories, performing dense conversion on the numerical vectors and the category vectors to obtain dense vectors, and determining recipe vectors corresponding to the recipe data.
According to the embodiment of the invention, the physical examination data and the recipe data are acquired so as to be used for recommending proper recipes according to the physical examination result. Wherein, the physical examination data refers to data containing numbers and types, such as height numbers and body fluid type physical examination data, and the recipe data refers to various types of food data, such as tomato and egg data.
Further, the embodiment of the invention is used for primarily classifying physical examination by identifying the physical examination values and the physical examination categories in the physical examination data, so that the data disorder in the subsequent data analysis is reduced. Wherein the physical examination value refers to a value in physical examination data, such as systolic pressure <120mmHg, diastolic pressure <80mmHg, and fasting blood glucose <5.6mmol/L, the physical examination category refers to a category of physical examination report, including a first category and a second category, for example, the first category includes a conventional physical examination (height, weight, blood pressure, internal medicine, surgery, otorhinolaryngology department, oral cavity, gynecology), blood examination (clinical examination, biochemical examination, immune examination, chemiluminescence and exemption examination, blood rheological examination), urine and secretion examination, B-ultrasonic, x-ray, CT, MR, and the like, and the second category refers to a category under the first category, such as height, weight, blood pressure, internal medicine, surgery, otorhinolaryngology department, oral cavity, gynecology, and the like when the first category is the conventional physical examination.
In an embodiment of the present invention, the identifying the physical examination number and the physical examination category in the physical examination data is implemented by identifying a physical examination item in the physical examination data.
For example, if the physical examination item is identified with a numerical value, the physical examination value is represented, the type of the identified numerical value is a height item, a secondary category may be determined, and if the physical examination item is identified with a clinical examination, a biochemical examination, an immunological examination, a chemiluminescent and exempt examination, and a blood rheological examination, a primary category may be determined.
Further, the embodiment of the invention constructs the numerical vector and the category vector of the physical examination numerical value and the physical examination category, so as to be used for converting the irregular physical examination data into the vector data in a unified format, thereby facilitating the subsequent analysis of the data by using a deep learning model. Wherein the numerical vector and the class vector refer to a vector of numerical representations between 0 and 1.
In an embodiment of the present invention, referring to fig. 2, the constructing a numeric vector and a category vector corresponding to the physical examination numeric value and the physical examination category includes:
s201, inquiring the numerical name of the physical examination numerical value;
s202, constructing a name vector and a category vector of the numerical name corresponding to the physical examination category by utilizing a single-hot encoding algorithm;
S203, calculating the numerical code of the physical examination numerical value by using the following formula:
wherein σ represents the numerical code of the physical examination value, and x represents the physical examination value;
s204, vector splicing is carried out on the numerical code and the name vector, and a numerical vector corresponding to the physical examination numerical value is obtained.
Further, the embodiment of the invention adapts to the structural characteristics of deep learning by performing dense conversion on the numerical vector and the category vector to be used for converting sparse vectors into dense vectors.
In an embodiment of the present invention, the performing a dense transformation on the numeric vector and the category vector to obtain a dense vector is implemented by using an embedding layer.
The embedded layer is an embedding layer, is a layer in a neural network structure, is composed of embedding_size neurons, is output of an input layer, and is used for mapping a high-dimensional sparse vector into a low-dimensional dense vector.
Further, the embodiment of the invention is convenient for the subsequent analysis of the data by using the deep learning model by determining the recipe vector corresponding to the recipe data and converting the irregular recipe data into the vector data in a unified format. Wherein, the recipe vector refers to a vector represented by 0 and 1.
In an embodiment of the present invention, the principle of determining the recipe vector corresponding to the recipe data is similar to the principle of constructing the category vector corresponding to the physical examination category by using the single thermal encoding algorithm, and will not be further described herein.
S2, calculating a correlation coefficient between the dense vector and the recipe vector, calculating a cross feature between the dense vector and the recipe vector according to the correlation coefficient, and determining a feature score corresponding to the cross feature.
According to the embodiment of the invention, the correlation coefficient between the dense vector and the recipe vector is calculated to be used for the correlation degree between the physical examination data and the recipe data, so that the recipe data suitable for the physical examination data can be determined by using the correlation degree.
In an embodiment of the present invention, the calculating the correlation coefficient between the dense vector and the recipe vector includes: calculating an association weight between the dense vector and the recipe vector using the formula:
wherein a is ij Representing the associated weights, dense representing the fully connected layer,
represents the dense vector, pe represents the physical examination data, T represents a transpose symbol, I re Representing the recipe vector, re representing the recipe data, i representing the serial number of the physical examination data, j representing the serial number of the recipe data;
determining an association coefficient between the dense vector and the recipe vector according to the association weight using the following formula:
wherein Weight is attention Representing the correlation coefficient between the dense vector and the recipe vector, the attention represents the attention mechanism layer, a 11 、a 1J 、a I1 、a IJ The association weight I, J indicates the number of serial numbers of the physical examination data and the number of serial numbers of the recipe data.
Further, according to the embodiment of the invention, the cross characteristic between the dense vector and the recipe vector is calculated according to the association coefficient so as to splice the dense vector and the recipe vector, and the relation between the dense vector and the recipe vector is mined, so that the matching of subsequent physical examination data and recipe data is ensured.
In an embodiment of the present invention, the calculating the cross feature between the dense vector and the recipe vector according to the association coefficient includes: calculating a dense cross feature of the dense vector from the correlation coefficient using the formula:
Wherein beta is i Representing a dense cross feature of the dense vector, J representing the number of sequence numbers of the recipe data, pe representing the physical examination data, i representing the sequence number of the physical examination data, J representing the sequence number of the recipe data,
representing an ith dense vector corresponding to the physical examination data;
according to the correlation coefficient, the recipe intersection characteristic of the recipe vector is calculated by using the following formula:
wherein alpha is j A recipe cross feature representing the recipe vector, I representing the number of serial numbers of the physical examination data, re representing the recipe data, I representing the serial number of the physical examination data, j representing the serial number of the recipe data,
a j-th recipe vector representing the recipe data;
calculating the intersection feature between the dense vector and the recipe vector from the dense intersection feature and the recipe intersection feature using the following formula:
wherein X is attention Representing the intersection features between the dense vector and the recipe vector, alpha representing the recipe intersection features of the recipe vector, beta representing the dense intersection features of the dense vector,
represents the dense vector, pe represents the physical examination data, T represents a transpose symbol, I re Representing the recipe vector, re representing the recipe data.
Further, the embodiment of the invention is used for determining the attention score corresponding to the cross feature by determining the feature score corresponding to the cross feature, namely the matching score of the physical examination data and the corresponding recipe data.
In one embodiment of the present invention, the feature score corresponding to the intersection feature is determined using the following formula:
X′ attention ＝w attention X attention +b attention
wherein X 'is' attention Representing the feature score, w, corresponding to the cross feature attention Network weights representing the attention mechanism layer, b attention Representing parameter deviation of the attention mechanism layer, X attention Representing a cross feature between the dense vector and the recipe vector.
And S3, constructing a feature cascade of the physical examination data by using the feature score and the dense vector, performing deep cross processing on the feature cascade to obtain a deep cross cascade, and calculating a recipe recommendation probability corresponding to the deep cross cascade.
The embodiment of the invention constructs the feature cascade of the physical examination data by utilizing the feature score and the dense vector, and is used for carrying out cross combination on the dense vector and the cross feature.
In an embodiment of the present invention, the constructing the feature cascade of the physical examination data using the feature score and the dense vector is implemented by using the following formula:
Wherein x is 0 Representing the cascade of features, X' attention Representing the feature score corresponding to the cross feature, U representing the dense vector, T representing the transposed symbol, emmbed, 1 representing X' attention And the row and column of U, k represents X' attention And the number of columns of U.
Further, the embodiment of the invention is used for improving the interaction degree between the dense vector and the recipe vector and improving the accuracy of matching the physical examination data with the recipe data by carrying out the deep cross processing on the feature cascade.
In an embodiment of the present invention, the performing a deep cross process on the feature cascade to obtain a deep cross cascade includes: performing multi-layer cross processing on the characteristic cascade by using the following formula to obtain a multi-layer cross cascade:
wherein X' attention Representing the multi-layer cross-cascade, x 0 Representing the cascade of features, w attention Network weights representing the attention mechanism layer, b attention Representing parameter deviation of the attention mechanism layer, X attention Representing a cross feature between the dense vector and the recipe vector, T representing a transpose symbol;
the feature depth combination of the feature cascade is constructed using the following formula:
h′ attention ＝f(w attention h attention +b attention )
wherein h' attention A feature depth combination, w, representing the feature cascade attention Network weights representing the attention mechanism layer, b attention Indicating parameter deviation of the attention mechanism layer, h attention Representing deep network layersF represents a function of the depth network layer;
and determining the depth cross cascade of the feature cascade according to the multi-layer cross cascade and the feature depth combination.
Wherein the depth cross cascade is composed of the multi-layer cross cascade and corresponding characteristic depth combinations thereof.
Further, the embodiment of the invention calculates the recipe recommendation probability corresponding to the depth cross cascade, so as to label the recipe data corresponding to the physical examination data by using the probability.
In an embodiment of the present invention, the recipe recommendation probability corresponding to the deep cross cascade is calculated using the following formula:
wherein p represents recipe recommendation probability corresponding to the deep cross cascade, X attention Representing the multi-layer Cross-cascade of the deep Cross-cascades, representing the output of the Cross Network, h' attention A feature depth combination representing the feature cascade in the depth cross cascade, an output representing Deep Network, w logits Representing the weights of the combination layer Combination Layer, the logits are derived by combining the outputs of the Cross Network and Deep Network through the combination layer Combination Layer and then summing them by a weight, σ representing the sigmoid function.
S4, determining a recipe recommendation result corresponding to the physical examination data in the recipe data according to the recipe recommendation probability.
According to the embodiment of the invention, the recipe recommendation result corresponding to the physical examination data is determined in the recipe data according to the recipe recommendation probability, so that the recommended recipe corresponding to the physical examination data is determined according to the output result of the model, intelligent decision is made by using the model, and the problem of inefficiency of manual recommendation is reduced. The recipe recommendation result is a recommendation result formed by combining recipes corresponding to the recipe recommendation probabilities.
In an embodiment of the present invention, referring to fig. 3, the determining, according to the recipe recommendation probability, a recipe recommendation result corresponding to the physical examination data in the recipe data includes:
s301, extracting a target recommendation probability which is larger than a preset probability from the recipe recommendation probabilities;
s302, inquiring target recipe data corresponding to the target recommendation probability in the recipe data;
s303, carrying out recipe data combination on the target recipe data to obtain a recipe recommendation result corresponding to the physical examination data.
It can be seen that, in the embodiment of the present invention, firstly, physical examination data and recipe data are obtained for recommending a proper recipe according to a physical examination result, further, in the embodiment of the present invention, physical examination values and physical examination categories in the physical examination data are identified for preliminary classification of physical examination, so as to reduce data disorder during subsequent data analysis, further, in the embodiment of the present invention, numerical vectors and category vectors corresponding to the physical examination values and the physical examination categories are constructed for converting irregular physical examination data into vector data in a uniform format, so that subsequent analysis of data by using a deep learning model is facilitated, further, in the embodiment of the present invention, by performing dense conversion on the numerical vectors and the category vectors, for converting sparse vectors into dense vectors, adapting to structural features of deep learning, further, the embodiment of the invention is convenient for the subsequent analysis of data by using a deep learning model by determining the recipe vector corresponding to the recipe data and converting the irregular recipe data into the vector data in a unified format, further, the embodiment of the invention can determine the recipe data suitable for physical examination data by calculating the association coefficient between the dense vector and the recipe vector and using the association degree by calculating the association coefficient between the physical examination data and the recipe vector, further, the embodiment of the invention can ensure the matching of the subsequent physical examination data and the recipe data by calculating the cross characteristic between the dense vector and the recipe vector according to the association coefficient, further, the embodiment of the invention constructs a feature cascade of the physical examination data by using the feature score and the dense vector to be used for carrying out cross combination on the dense vector and the cross feature by determining the feature score corresponding to the cross feature to be used for determining the attention score corresponding to the cross feature, namely the matching score of the physical examination data and the corresponding recipe data, further, the embodiment of the invention improves the accuracy of matching the physical examination data and the recipe data by carrying out deep cross processing on the feature cascade to be used for improving the interaction degree between the dense vector and the recipe vector, further, the embodiment of the invention improves the accuracy of matching the physical examination data and the recipe data by calculating the recipe recommendation probability corresponding to the deep cross cascade to be used for marking the recipe data corresponding to the physical examination data by using the probability, and further, the embodiment of the invention determines a recipe recommendation result corresponding to the physical examination data in the recipe data by using the recipe recommendation probability according to the recipe recommendation probability to be used for determining the recipe recommendation corresponding to the physical examination data according to the output result of the model, and further, the embodiment of the invention reduces the intelligent manpower problem by using the model. Therefore, the recipe intelligent recommendation method based on physical examination data can improve accuracy of user image classification.
FIG. 4 is a functional block diagram of the intelligent recipe recommendation device based on physical examination data.
The intelligent recipe recommendation device 400 based on physical examination data can be installed in electronic equipment. Depending on the implemented functions, the intelligent recipe recommendation device based on physical examination data may include a recipe vector determination module 401, a feature score determination module 402, a recommendation probability calculation module 403, and a recommendation result determination module 404. The module of the invention, which may also be referred to as a unit, refers to a series of computer program segments, which are stored in the memory of the electronic device, capable of being executed by the processor of the electronic device and of performing a fixed function.
In the embodiment of the present invention, the functions of each module/unit are as follows:
the recipe vector determining module 401 is configured to obtain physical examination data and recipe data, identify physical examination values and physical examination categories in the physical examination data, construct numerical vectors and category vectors corresponding to the physical examination values and the physical examination categories, perform dense conversion on the numerical vectors and the category vectors to obtain dense vectors, and determine recipe vectors corresponding to the recipe data;
The feature score determining module 402 is configured to calculate a correlation coefficient between the dense vector and the recipe vector, calculate a cross feature between the dense vector and the recipe vector according to the correlation coefficient, and determine a feature score corresponding to the cross feature;
the recommendation probability calculation module 403 is configured to construct a feature cascade of the physical examination data by using the feature score and the dense vector, perform deep cross processing on the feature cascade, obtain a deep cross cascade, and calculate a recipe recommendation probability corresponding to the deep cross cascade;
the recommendation result determining module 404 is configured to determine, according to the recipe recommendation probability, a recipe recommendation result corresponding to the physical examination data in the recipe data.
In detail, the modules in the intelligent recipe recommendation device 400 based on physical examination data in the embodiment of the present invention use the same technical means as the intelligent recipe recommendation method based on physical examination data described in fig. 1 to 3, and can generate the same technical effects, which are not described herein.
Fig. 5 is a schematic structural diagram of an electronic device for implementing the intelligent recipe recommendation method based on physical examination data.
The electronic device may comprise a processor 50, a memory 51, a communication bus 52 and a communication interface 53, and may further comprise a computer program stored in the memory 51 and executable on the processor 50, such as a recipe intelligent recommendation program based on physical examination data.
The processor 50 may be formed by an integrated circuit in some embodiments, for example, a single packaged integrated circuit, or may be formed by a plurality of integrated circuits packaged with the same function or different functions, including one or more central processing units (Central Processing unit, CPU), a microprocessor, a digital processing chip, a graphics processor, a combination of various control chips, and so on. The processor 50 is a Control Unit (Control Unit) of the electronic device, connects various components of the entire electronic device using various interfaces and lines, executes various functions of the electronic device and processes data by running or executing programs or modules stored in the memory 51 (for example, executing a recipe intelligent recommendation program based on physical examination data, etc.), and calling data stored in the memory 51.
The memory 51 includes at least one type of readable storage medium including flash memory, a removable hard disk, a multimedia card, a card type memory (e.g., SD or DX memory, etc.), a magnetic memory, a magnetic disk, an optical disk, etc. The memory 51 may in some embodiments be an internal storage unit of the electronic device, such as a mobile hard disk of the electronic device. The memory 51 may also be an external storage device of the electronic device in other embodiments, for example, a plug-in mobile hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card) or the like. Further, the memory 51 may also include both an internal storage unit and an external storage device of the electronic device. The memory 51 may be used not only for storing application software installed in an electronic device and various types of data, such as codes of a database-configured connection program, but also for temporarily storing data that has been output or is to be output.
The communication bus 52 may be a peripheral component interconnect standard (peripheral component interconnect, PCI) bus, or an extended industry standard architecture (extended industry standard architecture, EISA) bus, among others. The bus may be classified as an address bus, a data bus, a control bus, etc. The bus is arranged to enable a connection communication between the memory 51 and at least one processor 50 etc.
The communication interface 53 is used for communication between the electronic device 5 and other devices, including a network interface and a user interface. Optionally, the network interface may include a wired interface and/or a wireless interface (e.g., WI-FI interface, bluetooth interface, etc.), typically used to establish a communication connection between the electronic device and other electronic devices. The user interface may be a Display (Display), an input unit such as a Keyboard (Keyboard), or alternatively a standard wired interface, a wireless interface. Alternatively, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode) touch, or the like. The display may also be referred to as a display screen or display unit, as appropriate, for displaying information processed in the electronic device and for displaying a visual user interface.
Fig. 5 shows only an electronic device with components, and it will be understood by those skilled in the art that the structure shown in fig. 5 is not limiting of the electronic device and may include fewer or more components than shown, or may combine certain components, or a different arrangement of components.
For example, although not shown, the electronic device may further include a power source (such as a battery) for supplying power to the respective components, and the power source may be logically connected to the at least one processor 50 through a power management device, so that functions of charge management, discharge management, and power consumption management are implemented through the power management device. The power supply may also include one or more of any of a direct current or alternating current power supply, recharging device, power failure detection circuit, power converter or inverter, power status indicator, etc. The electronic device may further include various sensors, bluetooth modules, wi-Fi modules, etc., which are not described herein.
It should be understood that the embodiments described are for illustrative purposes only and are not limited in scope by this configuration.
The database-configured connection program stored in the memory 51 in the electronic device is a combination of a plurality of computer programs, which, when run in the processor 50, can implement:
Acquiring physical examination data and recipe data, identifying physical examination values and physical examination categories in the physical examination data, constructing numerical vectors and category vectors corresponding to the physical examination values and the physical examination categories, performing dense conversion on the numerical vectors and the category vectors to obtain dense vectors, and determining recipe vectors corresponding to the recipe data;
calculating a correlation coefficient between the dense vector and the recipe vector, calculating a cross feature between the dense vector and the recipe vector according to the correlation coefficient, and determining a feature score corresponding to the cross feature;
constructing a feature cascade of the physical examination data by using the feature score and the dense vector, performing deep cross processing on the feature cascade to obtain a deep cross cascade, and calculating a recipe recommendation probability corresponding to the deep cross cascade;
and determining a recipe recommendation result corresponding to the physical examination data in the recipe data according to the recipe recommendation probability.
In particular, the specific implementation method of the processor 50 on the computer program may refer to the description of the relevant steps in the corresponding embodiment of fig. 1, which is not repeated herein.
Further, the electronic device integrated modules/units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a non-volatile computer readable storage medium. The storage medium may be volatile or nonvolatile. For example, the computer readable medium may include: any entity or device capable of carrying the computer program code, a recording medium, a U disk, a removable hard disk, a magnetic disk, an optical disk, a computer Memory, a Read-Only Memory (ROM).
The present invention also provides a storage medium storing a computer program which, when executed by a processor of an electronic device, can implement:
acquiring physical examination data and recipe data, identifying physical examination values and physical examination categories in the physical examination data, constructing numerical vectors and category vectors corresponding to the physical examination values and the physical examination categories, performing dense conversion on the numerical vectors and the category vectors to obtain dense vectors, and determining recipe vectors corresponding to the recipe data;
calculating a correlation coefficient between the dense vector and the recipe vector, calculating a cross feature between the dense vector and the recipe vector according to the correlation coefficient, and determining a feature score corresponding to the cross feature;
constructing a feature cascade of the physical examination data by using the feature score and the dense vector, performing deep cross processing on the feature cascade to obtain a deep cross cascade, and calculating a recipe recommendation probability corresponding to the deep cross cascade;
and determining a recipe recommendation result corresponding to the physical examination data in the recipe data according to the recipe recommendation probability.
In the several embodiments provided in the present invention, it should be understood that the disclosed apparatus, device and method may be implemented in other manners. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the modules is merely a logical function division, and there may be other manners of division when actually implemented.
The modules described as separate components may or may not be physically separate, and components shown as modules may or may not be physical units, may be located in one place, or may be distributed over multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional module in the embodiments of the present invention may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units can be realized in a form of hardware or a form of hardware and a form of software functional modules.
It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof.
The present embodiments are, therefore, to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference signs in the claims shall not be construed as limiting the claim concerned.
It should be noted that in this document, relational terms such as "first" and "second" and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
The foregoing is only a specific embodiment of the invention to enable those skilled in the art to understand or practice the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (6)
1. An intelligent recipe recommendation method based on physical examination data is characterized by comprising the following steps:
acquiring physical examination data and recipe data, identifying physical examination values and physical examination categories in the physical examination data, respectively constructing numerical vectors and category vectors corresponding to the physical examination values and the physical examination categories, performing dense conversion on the numerical vectors and the category vectors to obtain dense vectors of the physical examination data, and determining recipe vectors corresponding to the recipe data;
calculating an association weight between the dense vector and the recipe vector using the formula:
wherein a is ij Representing the associated weights, dense representing the fully connected layer, 
Represents the dense vector, pe represents the physical examination data, T represents a transpose symbol, I re Representing the recipe vector, re representing the recipe data, i representing the serial number of the physical examination data, j representing the serial number of the recipe data,
according to the association weight, calculating a dense cross feature of the dense vector by using the following formula:
wherein beta is i Representing the dense cross feature of the dense vector, J representing the number of sequence numbers of the recipe data, pe representing the physical examination data, i representing the sequence number of the physical examination data, J representing the sequence number of the recipe dataThe number of the code is given, the code,
representing the ith dense vector, a, corresponding to the physical examination data ij Representing the weight of the association in question,
according to the association weight, the recipe intersection characteristic of the recipe vector is calculated by using the following formula:
wherein alpha is j A recipe cross feature representing the recipe vector, I representing the number of serial numbers of the physical examination data, re representing the recipe data, I representing the serial number of the physical examination data, j representing the serial number of the recipe data,
a j-th recipe vector corresponding to the recipe data, a ij Representing the weight of the association in question,
calculating the intersection feature between the dense vector and the recipe vector from the dense intersection feature and the recipe intersection feature using the following formula:
Wherein X is attention Representing the cross feature between the dense vector and the recipe vector, alpha j A recipe cross feature, beta, representing the recipe vector i Represents the dense intersection characteristics of the dense vector,
represents the dense vector, pe represents the physical examination data, T represents a transpose symbol, I re Representing the recipe vector, re representing the recipe data,
and determining feature scores corresponding to intersecting features between the dense vector and the recipe vector;
based on the feature score and the dense vector, constructing a feature cascade of the physical examination data using the following formula:
wherein x is 0 Representing the cascade of features, X' attention A feature score representing a cross feature correspondence between the dense vector and the recipe vector, U representing the dense vector, T representing a transposed symbol, and ebed representing (X' attention ) T And U T Is 1 (X' attention ) T And U T And k represents (X' attention ) T And U T Is used for the number of columns of (a),
performing multi-layer cross processing on the characteristic cascade by using the following formula to obtain a multi-layer cross cascade:
wherein X' attention Representing the multi-layer cross-cascade, x 0 Representing the cascade of features, w attention Network weights representing the attention mechanism layer of a crossover network, b attention Parameter bias, X, representing the attention mechanism layer of a crossover network attention Representing the cross-characteristics between the dense vector and the recipe vector, T representing the transposed symbol,
the feature depth combination of the feature cascade is constructed using the following formula:
h′ attention ＝f(w′ attention h attention +b′ attention )
wherein h' attention A feature depth combination representing the feature cascade, w' attention Attention machine for representing depth networkNetwork weight of the system layer, b' attention Parameter bias representing the attention mechanism layer of the depth network, h attention Representing the fully connected layer of the deep network layer, f represents a function of the deep network layer,
determining a depth cross cascade of the feature cascade according to the multi-layer cross cascade and the feature depth combination, and calculating a recipe recommendation probability corresponding to the depth cross cascade;
and determining a recipe recommendation result corresponding to the physical examination data in the recipe data according to the recipe recommendation probability.
2. The method of claim 1, wherein the constructing the numeric vector and the category vector of the physical examination numeric value and the physical examination category, respectively, comprises:
inquiring the numerical name of the physical examination numerical value;
constructing a name vector and a category vector corresponding to the numerical name and the physical examination category by using a single-hot encoding algorithm;
Calculating the numerical code of the physical examination value by using the following formula:
wherein σ represents the numerical code of the physical examination value, and x represents the physical examination value;
and vector splicing is carried out on the numerical code and the name vector, so that a numerical vector corresponding to the physical examination numerical value is obtained.
3. The method of claim 1, wherein determining a recipe recommendation corresponding to the physical examination data in the recipe data according to the recipe recommendation probability comprises:
extracting a target recommendation probability which is larger than a preset probability from the recipe recommendation probability;
inquiring target recipe data corresponding to the target recommendation probability in the recipe data;
and carrying out recipe data combination on the target recipe data to obtain a recipe recommendation result corresponding to the physical examination data.
4. An intelligent recipe recommendation device based on physical examination data, which is characterized by comprising:
the recipe vector determining module is used for acquiring physical examination data and recipe data, identifying physical examination values and physical examination categories in the physical examination data, respectively constructing numerical vectors and category vectors corresponding to the physical examination values and the physical examination categories, performing dense conversion on the numerical vectors and the category vectors to obtain dense vectors of the physical examination data, and determining recipe vectors corresponding to the recipe data;
A feature score determination module for calculating an association weight between the dense vector and the recipe vector using the following formula:
wherein a is ij Representing the associated weights, dense representing the fully connected layer,
represents the dense vector, pe represents the physical examination data, T represents a transpose symbol, I re Representing the recipe vector, re representing the recipe data, i representing the serial number of the physical examination data, j representing the serial number of the recipe data,
according to the association weight, calculating a dense cross feature of the dense vector by using the following formula:
wherein beta is i A dense cross feature representing the dense vector, J representing the recipeThe number of serial numbers of the data, pe represents the physical examination data, i represents the serial number of the physical examination data, j represents the serial number of the recipe data,
representing the ith dense vector, a, corresponding to the physical examination data ij Representing the weight of the association in question,
according to the association weight, the recipe intersection characteristic of the recipe vector is calculated by using the following formula:
wherein alpha is j A recipe cross feature representing the recipe vector, I representing the number of serial numbers of the physical examination data, re representing the recipe data, I representing the serial number of the physical examination data, j representing the serial number of the recipe data, 
A j-th recipe vector corresponding to the recipe data, a ij Representing the weight of the association in question,
calculating the intersection feature between the dense vector and the recipe vector from the dense intersection feature and the recipe intersection feature using the following formula:
wherein X is attention Representing the cross feature between the dense vector and the recipe vector, alpha j A recipe cross feature, beta, representing the recipe vector i Represents the dense intersection characteristics of the dense vector,
represents the dense vector, pe represents the physical examination data, T represents transposeSymbol I re Representing the recipe vector, re representing the recipe data,
and determining feature scores corresponding to intersecting features between the dense vector and the recipe vector;
the recommendation probability calculation module is used for constructing a feature cascade of the physical examination data based on the feature score and the dense vector by using the following formula:
wherein x is 0 Representing the cascade of features, X' attention A feature score representing a cross feature correspondence between the dense vector and the recipe vector, U representing the dense vector, T representing a transposed symbol, and ebed representing (X' attention ) T And U T Is 1 (X' attention ) T And U T And k represents (X' attention ) T And U T Is used for the number of columns of (a),
performing multi-layer cross processing on the characteristic cascade by using the following formula to obtain a multi-layer cross cascade:
wherein X' attention Representing the multi-layer cross-cascade, x 0 Representing the cascade of features, w attention Network weights representing the attention mechanism layer of a crossover network, b attention Parameter bias, X, representing the attention mechanism layer of a crossover network attention Representing the cross-characteristics between the dense vector and the recipe vector, T representing the transposed symbol,
the feature depth combination of the feature cascade is constructed using the following formula:
h′ attention ＝f(w′ attention h attention +b′ attention )
wherein h' attention A feature depth combination representing the feature cascade, w' attention Network weights, b ', representing the attention mechanism layer of a deep network' attention Parameter bias representing the attention mechanism layer of the depth network, h attention Representing the fully connected layer of the deep network layer, f represents a function of the deep network layer,
determining a depth cross cascade of the feature cascade according to the multi-layer cross cascade and the feature depth combination, and calculating a recipe recommendation probability corresponding to the depth cross cascade;
and the recommended result determining module is used for determining a recipe recommended result corresponding to the physical examination data in the recipe data according to the recipe recommended probability.
5. An electronic device, the electronic device comprising:
at least one processor; the method comprises the steps of,
a memory communicatively coupled to the at least one processor; wherein,,
the memory stores a computer program that is executed by the at least one processor to enable the at least one processor to perform the intelligent recipe recommendation method based on physical examination data as claimed in any one of claims 1 to 3.
6. A computer readable storage medium storing a computer program, wherein the computer program when executed by a processor implements the intelligent recipe recommendation method based on physical examination data as claimed in any one of claims 1 to 3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211260581.3A CN115547457B (en) | 2022-10-14 | 2022-10-14 | Recipe intelligent recommendation method, device, equipment and medium based on physical examination data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211260581.3A CN115547457B (en) | 2022-10-14 | 2022-10-14 | Recipe intelligent recommendation method, device, equipment and medium based on physical examination data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115547457A CN115547457A (en) | 2022-12-30 |
| CN115547457B true CN115547457B (en) | 2023-06-30 |
Family
ID=84734697
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211260581.3A Active CN115547457B (en) | 2022-10-14 | 2022-10-14 | Recipe intelligent recommendation method, device, equipment and medium based on physical examination data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115547457B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112420162A (en) * | 2020-11-10 | 2021-02-26 | 广州富港万嘉智能科技有限公司 | Intelligent recipe recommendation method and device and intelligent cabinet |
| CN113889209A (en) * | 2021-09-26 | 2022-01-04 | 浙江禾连网络科技有限公司 | Recommendation system and storage medium for health management service products |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108766528B (en) * | 2018-05-08 | 2021-07-16 | 天津大学 | Diet management system, construction method thereof and food material management method |
| CN109903836A (en) * | 2019-03-31 | 2019-06-18 | 山西慧虎健康科技有限公司 | A kind of diet intelligent recommendation and matching system and method based on constitution and big data |
| WO2022133190A1 (en) * | 2020-12-17 | 2022-06-23 | Trustees Of Tufts College | Food and nutrient estimation, dietary assessment, evaluation, prediction and management |
| CN113837844A (en) * | 2021-10-19 | 2021-12-24 | 北京邮电大学 | Multi-cascade downstream enterprise recommendation system and method and storage medium |
-
2022
- 2022-10-14 CN CN202211260581.3A patent/CN115547457B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112420162A (en) * | 2020-11-10 | 2021-02-26 | 广州富港万嘉智能科技有限公司 | Intelligent recipe recommendation method and device and intelligent cabinet |
| CN113889209A (en) * | 2021-09-26 | 2022-01-04 | 浙江禾连网络科技有限公司 | Recommendation system and storage medium for health management service products |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115547457A (en) | 2022-12-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111814962B (en) | Parameter acquisition method and device for identification model, electronic equipment and storage medium | |
| WO2023015935A1 (en) | Method and apparatus for recommending physical examination item, device and medium | |
| CN113378970B (en) | Sentence similarity detection method and device, electronic equipment and storage medium | |
| CN114822812A (en) | Character dialogue simulation method, device, equipment and storage medium | |
| WO2023029511A1 (en) | Intelligent inspection method and apparatus for online medical prescription medication, and device and storage medium | |
| CN110867231A (en) | Disease prediction method, device, computer equipment and medium based on text classification | |
| CN113779179B (en) | ICD intelligent coding method based on deep learning and knowledge graph | |
| CN113158676A (en) | Professional entity and relationship combined extraction method and system and electronic equipment | |
| CN115238670B (en) | Information text extraction method, device, equipment and storage medium | |
| CN112071429A (en) | Medical automatic question-answering system construction method based on knowledge graph | |
| CN116821373A (en) | Map-based prompt recommendation method, device, equipment and medium | |
| CN114840684A (en) | Map construction method, device and equipment based on medical entity and storage medium | |
| CN113157739A (en) | Cross-modal retrieval method and device, electronic equipment and storage medium | |
| CN115547457B (en) | Recipe intelligent recommendation method, device, equipment and medium based on physical examination data | |
| CN116844711A (en) | Disease auxiliary identification method and device based on deep learning | |
| CN116884612A (en) | Intelligent analysis method, device, equipment and storage medium for disease risk level | |
| CN116720525A (en) | Disease auxiliary analysis method, device, equipment and medium based on inquiry data | |
| CN114596958B (en) | Pathological data classification method, device, equipment and medium based on cascade classification | |
| CN116719904A (en) | Information query method, device, equipment and storage medium based on image-text combination | |
| CN116741358A (en) | Inquiry registration recommendation method, inquiry registration recommendation device, inquiry registration recommendation equipment and storage medium | |
| CN116702776A (en) | Multi-task semantic division method, device, equipment and medium based on cross-Chinese and western medicine | |
| CN116665878A (en) | Intelligent inquiry method, device, equipment and storage medium for improving accumulated errors | |
| CN114511569A (en) | Tumor marker-based medical image identification method, device, equipment and medium | |
| CN117116384B (en) | Targeted induction type medical molecular structure generation method | |
| CN114864032B (en) | Clinical data acquisition method and device based on HIS system |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |