CN109783859A - Model building method, device and computer readable storage medium - Google Patents
Model building method, device and computer readable storage medium Download PDFInfo
- Publication number
- CN109783859A CN109783859A CN201811524934.XA CN201811524934A CN109783859A CN 109783859 A CN109783859 A CN 109783859A CN 201811524934 A CN201811524934 A CN 201811524934A CN 109783859 A CN109783859 A CN 109783859A
- Authority
- CN
- China
- Prior art keywords
- model
- user
- data
- functionality controls
- column
- 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
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The present invention relates to intelligent decisions, disclose a kind of model building method, this method comprises: showing pre-set control selection interface when receiving the modeling instruction of user terminal transmission;Determine that user executes sequence from the functionality controls that the control selection region in control selection interface is that each modeling procedure selects, and between multiple functionality controls of determining user setting, wherein a modeling procedure corresponds to one or more functionality controls;Code corresponding with the functionality controls of user's selection is transferred from database, and is sequentially generated modeling procedure according to the execution between functionality controls;When receiving the operational order based on the functionality controls triggering in modeling procedure, the corresponding code of the functionality controls is executed, until completing the training of disaggregated model or regression model, and determines model parameter.The present invention also proposes a kind of model construction device and a kind of computer readable storage medium.The present invention improves modeling efficiency.
Description
Technical field
The present invention relates to intelligent Decision Technology field more particularly to a kind of model building methods, device and computer-readable
Storage medium.
Background technique
The modeling procedure that existing machine learning Modeling Platform provides is to predefine well, i.e., developer is according to preparatory
The modeling procedure determined carries out the exploitation of code, to construct specific modeling procedure.This mode causes user can not basis
The step of preference pattern training is needed, it, can only be corresponding by developing modeling procedure again if user wants change modeling procedure
The mode of code construct new modeling procedure, cause modeling efficiency low.
Summary of the invention
The present invention provides a kind of model building method, device and computer readable storage medium, main purpose and is to mention
High modeling efficiency.
To achieve the above object, the present invention also provides a kind of model building methods, this method comprises:
When receiving the modeling instruction of user terminal transmission, pre-set control selection interface is shown, wherein described
It include the control selection region of multiple modeling procedures in control selection interface, the multiple modeling procedure is for creating classification mould
Type or regression model;
Determine user from the function control that the control selection region in the control selection interface is that each modeling procedure selects
Part, and sequence is executed between the multiple functionality controls of determining user setting, wherein a modeling procedure corresponds to one
Or multiple functionality controls;
Code corresponding with the functionality controls of user's selection is transferred from database, and according to holding between the functionality controls
Row is sequentially generated modeling procedure;
When receiving the operational order based on the functionality controls triggering in the modeling procedure, the functionality controls pair are executed
The code answered until completing the training of disaggregated model or regression model, and determines model parameter.
Optionally, the modeling procedure successively includes data load step, data initialization step, data prediction step
Suddenly, characteristic processing step and model training step;It is described to receive based on the default step triggering in the modeling procedure
When operational order, the default corresponding code of step is executed, until completing the training of disaggregated model or regression model, and determines mould
The step of shape parameter includes:
When receiving the data load instruction that user is triggered based on the functionality controls of data load step, user's base is obtained
In the data file that data load step uploads, loads and show the data file;
After receiving the data initialization instruction that user is triggered based on the functionality controls of data initialization step, institute is parsed
Data file is stated, and determines the input feature vector column in the data file and target signature column;
When receiving the data prediction instruction that user is triggered based on the functionality controls of data prediction step, according to pre-
If preprocessing algorithms to the input feature vector column and the target signature column in feature pre-process;
When receiving the characteristic processing instruction that user is triggered based on the functionality controls of characteristic processing step, according to pretreatment
The weight of the corresponding feature of each characteristic series of input feature vector column count afterwards, and show the weight, so that user is according to displaying
Weight feature is filtered, and determine user selection input feature vector column;
When receiving the model training instruction that user is triggered based on the functionality controls of model training step, determine that user selects
The disaggregated model or regression model selected, the input feature vector column selected using user and the target signature arrange the training classification mould
Type or regression model are to determine model parameter.
Optionally, the parsing data file, and determine that the input feature vector column in the data file and target are special
Levying the step of arranging includes:
Identify the data type of the characteristic series and characteristic series in the data file, and based on the characteristic series and spy recognized
Levy data type display data role's set interface of column;
Determine user's input feature vector column that role's set interface is selected from data column based on the data and target
Characteristic series.
Optionally, if the disaggregated model of user's selection has the disaggregated model multiple, the determining user selects, use is used
The input feature vector column and the target signature column training disaggregated model or regression model of family selection are to determine model parameter
Step includes:
It determines multiple disaggregated models of user's selection, is arranged and the target signature using the input feature vector that user selects respectively
The multiple disaggregated model of column training, with the model parameter of each disaggregated model of determination;
After the step of model parameter of each disaggregated model of determination, the method also includes steps:
The accuracy rate and classification report of each disaggregated model are shown, so that user is according to the accuracy rate and classification report of displaying
Select optimal classification model;
The disaggregated model for determining user's selection, using the disaggregated model as object-class model.
Optionally, the modeling procedure further includes visualizing step, the parsing data file, and determines institute
After the step of stating input feature vector column and the target signature column in data file, the method also includes steps:
When receiving user based on the visual presentation instruction for visualizing the corresponding functionality controls creation of step, really
Determine the visual presentation mode of user setting;
The correlation for carrying out variable distribution to current data column is analyzed, and by analysis result according to user setting
Visual presentation mode is shown.
In addition, to achieve the above object, the present invention also provides a kind of model construction device, which includes memory and place
Device is managed, is stored with the model construction program that can be run on the processor, the model construction program quilt in the memory
The processor realizes following steps when executing:
When receiving the modeling instruction of user terminal transmission, pre-set control selection interface is shown, wherein described
It include the control selection region of multiple modeling procedures in control selection interface, the multiple modeling procedure is for creating classification mould
Type or regression model;
Determine user from the function control that the control selection region in the control selection interface is that each modeling procedure selects
Part, and sequence is executed between the multiple functionality controls of determining user setting, wherein a modeling procedure corresponds to one
Or multiple functionality controls;
Code corresponding with the functionality controls of user's selection is transferred from database, and according to holding between the functionality controls
Row is sequentially generated modeling procedure;
When receiving the operational order based on the functionality controls triggering in the modeling procedure, the functionality controls pair are executed
The code answered until completing the training of disaggregated model or regression model, and determines model parameter.
Optionally, the modeling procedure successively includes data load step, data initialization step, data prediction step
Suddenly, characteristic processing step and model training step;It is described to receive based on the default step triggering in the modeling procedure
When operational order, the default corresponding code of step is executed, until completing the training of disaggregated model or regression model, and determines mould
The step of shape parameter includes:
When receiving the data load instruction that user is triggered based on the functionality controls of data load step, user's base is obtained
In the data file that data load step uploads, loads and show the data file;
After receiving the data initialization instruction that user is triggered based on the functionality controls of data initialization step, institute is parsed
Data file is stated, and determines the input feature vector column in the data file and target signature column;
When receiving the data prediction instruction that user is triggered based on the functionality controls of data prediction step, according to pre-
If preprocessing algorithms to the input feature vector column and the target signature column in feature pre-process;
When receiving the characteristic processing instruction that user is triggered based on the functionality controls of characteristic processing step, according to pretreatment
The weight of the corresponding feature of each characteristic series of input feature vector column count afterwards, and show the weight, so that user is according to displaying
Weight feature is filtered, and determine user selection input feature vector column;
When receiving the model training instruction that user is triggered based on the functionality controls of model training step, determine that user selects
The disaggregated model or regression model selected, the input feature vector column selected using user and the target signature arrange the training classification mould
Type or regression model are to determine model parameter.
Optionally, the parsing data file, and determine that the input feature vector column in the data file and target are special
Levying the step of arranging includes:
Identify the data type of the characteristic series and characteristic series in the data file, and based on the characteristic series and spy recognized
Levy data type display data role's set interface of column;
Determine user's input feature vector column that role's set interface is selected from data column based on the data and target
Characteristic series.
Optionally, if the disaggregated model of user's selection has the disaggregated model multiple, the determining user selects, use is used
The input feature vector column and the target signature column training disaggregated model or regression model of family selection are to determine model parameter
Step includes:
It determines multiple disaggregated models of user's selection, is arranged and the target signature using the input feature vector that user selects respectively
The multiple disaggregated model of column training, with the model parameter of each disaggregated model of determination;
The model construction program can also be executed by the processor, to join in the model of each disaggregated model of the determination
After several steps, following steps are also realized:
The accuracy rate and classification report of each disaggregated model are shown, so that user is according to the accuracy rate and classification report of displaying
Select optimal classification model;
The disaggregated model for determining user's selection, using the disaggregated model as object-class model.
In addition, to achieve the above object, it is described computer-readable the present invention also provides a kind of computer readable storage medium
Model construction program is stored on storage medium, the model construction program can be executed by one or more processor, with reality
Now the step of model building method as described above.
Model building method, device and computer readable storage medium proposed by the present invention are sent out when receiving user terminal
When the modeling instruction sent, pre-set control selection interface is shown, wherein include multiple modeling steps in control selection interface
Rapid control selection region, multiple modeling procedures are for creating disaggregated model or regression model;Determine that user selects boundary from control
Control selection region on face is the functionality controls of each modeling procedure selection, and determine user setting multiple functionality controls it
Between execute sequence, wherein modeling procedure corresponds to one or more functionality controls;It is transferred and function control from database
The corresponding code of part, and modeling procedure is sequentially generated according to the execution between functionality controls;It is based in modeling procedure when receiving
Functionality controls triggering operational order when, execute the corresponding code of the functionality controls, until complete disaggregated model or return mould
The training of type, and determine model parameter.The program pre-sets each functionality controls that modeling procedure may be used, Yong Huke
To select arbitrary functionality controls creation modeling procedure as needed, even if after model construction completion, if user needs to become
More modeling procedure only needs replacement function control, without exploitation code again, improves modeling efficiency.
Detailed description of the invention
Fig. 1 is the flow diagram for the model building method that one embodiment of the invention provides;
Fig. 2 is the schematic diagram of internal structure for the model construction device that one embodiment of the invention provides;
The module diagram of model construction program in the model construction device that Fig. 3 provides for one embodiment of the invention.
The embodiments will be further described with reference to the accompanying drawings for the realization, the function and the advantages of the object of the present invention.
Specific embodiment
It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, it is not intended to limit the present invention.
The present invention provides a kind of model building method.It is the model construction that one embodiment of the invention provides shown in referring to Fig.1
The flow diagram of method.This method can be executed by a device, which can be by software and or hardware realization.
In the present embodiment, model building method includes:
Step S10 shows pre-set control selection interface when receiving the modeling instruction of user terminal transmission,
It wherein, include the control selection region of multiple modeling procedures in the control selection interface, the multiple modeling procedure is used for
Create disaggregated model or regression model.
Step S20, determine user from the control selection region in the control selection interface be each modeling procedure selection
Functionality controls, and determine user setting the multiple functionality controls between execute sequence, wherein a modeling procedure pair
It should be in one or more functionality controls.
Step S30 transfers code corresponding with the functionality controls of user's selection from database, and according to the functionality controls
Between execution be sequentially generated modeling procedure.
This programme provides a kind of general modeling, and user can according to need each modeling in setting modeling procedure
The specific algorithm of step, the system are provided with control selection interface, also referred to as workspace interface in advance, in the control selection interface
It include the control selection region of multiple modeling procedures, the multiple modeling procedure is for creating disaggregated model or regression model.
In control selection interface, a modeling procedure has corresponding multiple preset controls available, each preset control corresponding one
Kind data processing method.For example, data initialization module includes picture initialization control, sequence initialization control etc.;Data add
Carrying step includes data load control;Data prediction step includes standardizing control, continuously leaving scattered control etc..Specifically,
The system can be used for the modeling of disaggregated model or regression model, and for both models, which is provided with data load
Step, data initialization step, data prediction step, characteristic processing step and model training step, user as needed from
One or more preset control is selected in each step, and establishes the sequence that executes between preset control in sequence, is formed
One complete modeling procedure.
Specifically, user can select the preset control needed from control selection interface, and be dragged to modeling procedure
Editing area, establish the connection relationship between functionality controls, system determines the execution of each functionality controls according to connection relationship
Sequentially.Also, user can trigger the operational order to the functionality controls by click function control, to run the functionality controls
Corresponding code handles the data of input, and realizes the corresponding function of functionality controls.
In addition, it should be noted that, during executing the modeling procedure, the corresponding function of a upper modeling procedure
The output data of control is the input data of the corresponding functionality controls of next modeling procedure, the input number of first functionality controls
According to the data file uploaded for user.User can prepare the data file for being used on training pattern in advance, be loaded by data
Data file is uploaded to system as the data for being used for training pattern by step.
Step S40 executes the function when receiving the operational order based on the functionality controls triggering in the modeling procedure
The corresponding code of energy control until completing the training of disaggregated model or regression model, and determines model parameter.
In one embodiment, the multiple modeling procedure successively includes data load step, data initialization step, number
Data preprocess step, characteristic processing step and model training step;It is described to receive based on default in the modeling procedure
When the operational order of step triggering, the default corresponding code of step is executed, until completing the instruction of disaggregated model or regression model
Practice, and the step of determining model parameter includes:
When receiving the data load instruction that user is triggered based on the functionality controls of data load step, user's base is obtained
In the data file that data load step uploads, loads and show the data file;
After receiving the data initialization instruction that user is triggered based on the functionality controls of data initialization step, institute is parsed
Data file is stated, and determines the input feature vector column in the data file and target signature column.It can also be further to feature
Column and target column carry out Missing Data Filling, lack fill method there are many, including but not limited to following method: mode filling, in
Filling power is filled and specified manually to digit filling, mean value filling, maximum value filling, minimum value.
When receiving the data prediction instruction that user is triggered based on the functionality controls of data prediction step, according to pre-
If preprocessing algorithms to the input feature vector column and the target signature column in feature pre-process;
When receiving the characteristic processing instruction that user is triggered based on the functionality controls of characteristic processing step, according to pretreatment
The weight of the corresponding feature of each characteristic series of input feature vector column count afterwards, and show the weight, so that user is according to displaying
Weight feature is filtered, and determine user selection input feature vector column;
When receiving the model training instruction that user is triggered based on the functionality controls of model training step, determine that user selects
The disaggregated model or regression model selected, the input feature vector column selected using user and the target signature are arranged, the training classification
Model or regression model are to determine model parameter.
This programme is illustrated with a specific application scenarios below: by the general modeling system of this programme to state
There is the risk class of enterprise to be identified, user prepares the data file for being used on model training in advance, main in the data file
It to include essential information, trade information, the bankruptcy information of state-owned enterprise over the years, the policy factor etc. of state-owned enterprise, wherein basic
Information includes asset-liabilities information, cash stream information, Operating profit information, equity investment information, budget information, Project in Operation letter
Breath, crew size, worker's stability, Dong supervise high separation rate, recently the rewards and punishments information in year, recently whether voluntary bankruptcy, whether have
Innovation item etc..Wherein, the file format of data file can be csv, jpg, png, txt etc., wherein every in data file
By separators between a data column, and since the separator that the data file of different-format uses is different, it uses
Family needs to select corresponding separator in systems when uploading data file.System is when identifying data file, according to user
Data column in the separator identification data file of selection, if the separator in the separator and data file of user's selection is different
It causes, then will lead to the data column that can not normally identify in data file.
System carries out the displaying of data column after recognizing data column, so that setting data arrange user on showing interface
Data role, whole data column are divided into input feature vector column and target signature arranges.For the risk class of state-owned enterprise
For identification, in the data file of upload, the risk class of state-owned enterprise is target signature column, and other data are classified as input
Characteristic series.
Further, the parsing data file, and determine input feature vector column and target in the data file
The step of characteristic series includes:
Identify the data type of the characteristic series and characteristic series in the data file, and based on the characteristic series and spy recognized
Levy data type display data role's set interface of column;
Determine user's input feature vector column that role's set interface is selected from data column based on the data and target
Characteristic series.
Further, if the disaggregated model of user's selection has multiple, the determining user selects disaggregated model or return
Return model, the input feature vector column selected using user and the target signature are arranged, the training disaggregated model or regression model with
The step of determining model parameter include:
It determines multiple disaggregated models of user's selection, is arranged and the target signature using the input feature vector that user selects respectively
The multiple disaggregated model of column training, with the model parameter of each disaggregated model of determination;
The accuracy rate and classification report of each disaggregated model are shown, so that user is according to the accuracy rate and classification report of displaying
Select optimal classification model;
The disaggregated model for determining user's selection, using the disaggregated model as object-class model.
Disaggregated model in the program includes categorised decision tree-model, Random Forest model, neural network model, Bayes
Classifier and SVM (Support Vector Machine, support vector machines) classifier etc., regression model include returning to determine
Plan tree-model, linear regression model (LRM) etc..
Further, which is additionally provided with visual presentation step, and this method further includes following steps:
When receiving user based on the visual presentation instruction for visualizing the corresponding preset control creation of step, really
Determine the visual presentation mode of user setting;
The correlation for carrying out variable distribution to current data column is analyzed, and by analysis result according to user setting
Visual presentation mode is shown.Wherein, visual presentation mode includes box figure, pie chart, line chart, histogram, Q-Q figure (Q generation
Table quantile) etc., the type that user can according to need the data of displaying selects corresponding visual presentation mode.
Further, in one embodiment, a data file declustering can also be by the program using clustering algorithm
Multiple data subfiles.Then it is directed to different regression model or disaggregated model, multiple modeling procedures are constructed, by data Ziwen
Part is trained model respectively as the input data of multiple modeling procedures.Wherein, the classification in above-mentioned fractionation data file
Target data column in algorithm, with regression model or disaggregated model are not identical.For example, executing step S10 to step S30, structure
File declustering process is built, the data of input are split as multiple data subfiles by sorting algorithm, respectively obtains fractionation
Multiple data subfiles are inputted as data, execute step S10 to step S40 again.
The model building method that the present embodiment proposes is shown preparatory when receiving the modeling instruction of user terminal transmission
The control selection interface of setting, wherein include the control selection region of multiple modeling procedures, Duo Gejian in control selection interface
Mould step is for creating disaggregated model or regression model;Determine user from the control selection region in control selection interface be it is each
The functionality controls of modeling procedure selection, and sequence is executed between multiple functionality controls of determining user setting, wherein one is built
Mould step corresponds to one or more functionality controls;Code corresponding with functionality controls is transferred from database, and according to function
Execution between control is sequentially generated modeling procedure;When receive based in modeling procedure functionality controls triggering operational order
When, the corresponding code of the functionality controls is executed, until completing the training of disaggregated model or regression model, and determines model parameter.
The program pre-sets each functionality controls that modeling procedure may be used, and user can according to need the arbitrary function of selection
Control creates modeling procedure, even if, if user needs to change modeling procedure, only needing replacement function control after model construction completion
Part improves modeling efficiency without exploitation code again.
The present invention also provides a kind of model construction devices.It is the model structure that one embodiment of the invention provides referring to shown in Fig. 2
Build the schematic diagram of internal structure of device.
In the present embodiment, model construction device 1 can be PC (Personal Computer, PC), can also be with
It is the terminal devices such as smart phone, tablet computer, portable computer.The model construction device 1 includes at least memory 11, processing
Device 12, network interface 13 and communication bus.
Wherein, memory 11 include at least a type of readable storage medium storing program for executing, the readable storage medium storing program for executing include flash memory,
Hard disk, multimedia card, card-type memory (for example, SD or DX memory etc.), magnetic storage, disk, CD etc..Memory 11
It can be the internal storage unit of model construction device 1, such as the hard disk of the model construction device 1 in some embodiments.It deposits
Reservoir 11 is also possible in further embodiments on the External memory equipment of model construction device 1, such as model construction device 1
The plug-in type hard disk of outfit, intelligent memory card (Smart Media Card, SMC), secure digital (Secure Digital, SD)
Card, flash card (Flash Card) etc..Further, memory 11 can also both include the storage inside of model construction device 1
Unit also includes External memory equipment.Memory 11 can be not only used for the application software that storage is installed on model construction device 1
And Various types of data, such as the code of model construction program 01 etc., it can be also used for temporarily storing and exported or will be defeated
Data out.
Processor 12 can be in some embodiments a central processing unit (Central Processing Unit,
CPU), controller, microcontroller, microprocessor or other data processing chips, the program for being stored in run memory 11
Code or processing data, such as execute model construction program 01 etc..
Network interface 13 optionally may include standard wireline interface and wireless interface (such as WI-FI interface), be commonly used in
Communication connection is established between the device 1 and other electronic equipments.
Communication bus is for realizing the connection communication between these components.
Optionally, which can also include user interface, and user interface may include display (Display), input
Unit such as keyboard (Keyboard), optional user interface can also include standard wireline interface and wireless interface.It is optional
Ground, in some embodiments, display can be light-emitting diode display, liquid crystal display, touch-control liquid crystal display and OLED
(Organic Light-Emitting Diode, Organic Light Emitting Diode) touches device etc..Wherein, display can also be appropriate
Referred to as display screen or display unit, for being shown in the information handled in model construction device 1 and for showing visually
User interface.
Fig. 2 illustrates only the model construction device 1 with component 11-13 and model construction program 01, art technology
Personnel may include than illustrating more it is understood that structure shown in fig. 1 does not constitute the restriction to model construction device 1
Perhaps more component perhaps combines certain components or different component layouts less.
Optionally, which can also include touch sensor.It is touched provided by the touch sensor for user
The region for touching operation is known as touch area.In addition, touch sensor described here can be resistive touch sensor, capacitor
Formula touch sensor etc..Moreover, the touch sensor not only includes the touch sensor of contact, proximity may also comprise
Touch sensor etc..In addition, the touch sensor can be single sensor, or multiple sensings of array arrangement
Device.The area of the display of the device 1 can be identical as the area of the touch sensor, can also be different.Optionally, it will show
Show that device and touch sensor stacking are arranged, to form touch display screen.The device 1 is based on touch display screen detecting user's touching
The touch control operation of hair.
Optionally, which can also include camera, RF (Radio Frequency, radio frequency) circuit, sensor, sound
Frequency circuit, WiFi module etc..Wherein, sensor such as optical sensor, motion sensor and other sensors.Specifically, light
Sensor may include ambient light sensor and proximity sensor, wherein if the device 1 is mobile terminal, ambient light sensor can
The brightness of display screen is adjusted according to the light and shade of ambient light, proximity sensor can be closed when mobile terminal is moved in one's ear
Display screen and/or backlight.As a kind of motion sensor, gravity accelerometer can detect in all directions (generally
Three axis) acceleration size, can detect that size and the direction of gravity when static, can be used to identify the application of mobile terminal posture
(such as horizontal/vertical screen switching, dependent game, magnetometer pose calibrating), Vibration identification correlation function (such as pedometer, percussion) etc.;
Certainly, mobile terminal can also configure the other sensors such as gyroscope, barometer, hygrometer, thermometer, infrared sensor,
This is repeated no more.
In 1 embodiment of device shown in Fig. 2, model construction program 01 is stored in memory 11;Processor 12 executes
Following steps are realized when the model construction program 01 stored in memory 11:
When receiving the modeling instruction of user terminal transmission, pre-set control selection interface is shown, wherein described
It include the control selection region of multiple modeling procedures in control selection interface, the multiple modeling procedure is for creating classification mould
Type or regression model.
Determine user from the function control that the control selection region in the control selection interface is that each modeling procedure selects
Part, and sequence is executed between the multiple functionality controls of determining user setting, wherein a modeling procedure corresponds to one
Or multiple functionality controls.
Code corresponding with the functionality controls of user's selection is transferred from database, and according to holding between the functionality controls
Row is sequentially generated modeling procedure.
This programme provides a kind of general modeling, and user can according to need each modeling in setting modeling procedure
The specific algorithm of step, the system are provided with control selection interface, also referred to as workspace interface in advance, in the control selection interface
It include the control selection region of multiple modeling procedures, the multiple modeling procedure is for creating disaggregated model or regression model.
In control selection interface, a modeling procedure has corresponding multiple preset controls available, each preset control corresponding one
Kind data processing method.For example, data initialization module includes picture initialization control, sequence initialization control etc.;Data add
Carrying step includes data load control;Data prediction step includes standardizing control, continuously leaving scattered control etc..Specifically,
The system can be used for the modeling of disaggregated model or regression model, and for both models, which is provided with data load
Step, data initialization step, data prediction step, characteristic processing step and model training step, user as needed from
One or more preset control is selected in each step, and establishes the sequence that executes between preset control in sequence, is formed
One complete modeling procedure.
Specifically, user can select the preset control needed from control selection interface, and be dragged to modeling procedure
Editing area, establish the connection relationship between functionality controls, system determines the execution of each functionality controls according to connection relationship
Sequentially.Also, user can trigger the operational order to the functionality controls by click function control, to run the functionality controls
Corresponding code handles the data of input, and realizes the corresponding function of functionality controls.
In addition, it should be noted that, during executing the modeling procedure, the corresponding function of a upper modeling procedure
The output data of control is the input data of the corresponding functionality controls of next modeling procedure, the input number of first functionality controls
According to the data file uploaded for user.User can prepare the data file for being used on training pattern in advance, be loaded by data
Data file is uploaded to system as the data for being used for training pattern by step.
When receiving the operational order based on the functionality controls triggering in the modeling procedure, the functionality controls pair are executed
The code answered until completing the training of disaggregated model or regression model, and determines model parameter.
In one embodiment, the multiple modeling procedure successively includes data load step, data initialization step, number
Data preprocess step, characteristic processing step and model training step;It is described to receive based on default in the modeling procedure
When the operational order of step triggering, the default corresponding code of step is executed, until completing the instruction of disaggregated model or regression model
Practice, and the step of determining model parameter includes:
When receiving the data load instruction that user is triggered based on the functionality controls of data load step, user's base is obtained
In the data file that data load step uploads, loads and show the data file;
After receiving the data initialization instruction that user is triggered based on the functionality controls of data initialization step, institute is parsed
Data file is stated, and determines the input feature vector column in the data file and target signature column.It can also be further to feature
Column and target column carry out Missing Data Filling, lack fill method there are many, including but not limited to following method: mode filling, in
Filling power is filled and specified manually to digit filling, mean value filling, maximum value filling, minimum value.
When receiving the data prediction instruction that user is triggered based on the functionality controls of data prediction step, according to pre-
If preprocessing algorithms to the input feature vector column and the target signature column in feature pre-process;
When receiving the characteristic processing instruction that user is triggered based on the functionality controls of characteristic processing step, according to pretreatment
The weight of the corresponding feature of each characteristic series of input feature vector column count afterwards, and show the weight, so that user is according to displaying
Weight feature is filtered, and determine user selection input feature vector column;
When receiving the model training instruction that user is triggered based on the functionality controls of model training step, determine that user selects
The disaggregated model or regression model selected, the input feature vector column selected using user and the target signature are arranged, the training classification
Model or regression model are to determine model parameter.
This programme is illustrated with a specific application scenarios below: by the general modeling system of this programme to state
There is the risk class of enterprise to be identified, user prepares the data file for being used on model training in advance, main in the data file
It to include essential information, trade information, the bankruptcy information of state-owned enterprise over the years, the policy factor etc. of state-owned enterprise, wherein basic
Information includes asset-liabilities information, cash stream information, Operating profit information, equity investment information, budget information, Project in Operation letter
Breath, crew size, worker's stability, Dong supervise high separation rate, recently the rewards and punishments information in year, recently whether voluntary bankruptcy, whether have
Innovation item etc..Wherein, the file format of data file can be csv, jpg, png, txt etc., wherein every in data file
By separators between a data column, and since the separator that the data file of different-format uses is different, it uses
Family needs to select corresponding separator in systems when uploading data file.System is when identifying data file, according to user
Data column in the separator identification data file of selection, if the separator in the separator and data file of user's selection is different
It causes, then will lead to the data column that can not normally identify in data file.
System carries out the displaying of data column after recognizing data column, so that setting data arrange user on showing interface
Data role, whole data column are divided into input feature vector column and target signature arranges.For the risk class of state-owned enterprise
For identification, in the data file of upload, the risk class of state-owned enterprise is target signature column, and other data are classified as input
Characteristic series.
Further, the parsing data file, and determine input feature vector column and target in the data file
The step of characteristic series includes:
Identify the data type of the characteristic series and characteristic series in the data file, and based on the characteristic series and spy recognized
Levy data type display data role's set interface of column;
Determine user's input feature vector column that role's set interface is selected from data column based on the data and target
Characteristic series.
Further, if the disaggregated model of user's selection has multiple, the determining user selects disaggregated model or return
Return model, the input feature vector column selected using user and the target signature are arranged, the training disaggregated model or regression model with
The step of determining model parameter include:
It determines multiple disaggregated models of user's selection, is arranged and the target signature using the input feature vector that user selects respectively
The multiple disaggregated model of column training, with the model parameter of each disaggregated model of determination;
The accuracy rate and classification report of each disaggregated model are shown, so that user is according to the accuracy rate and classification report of displaying
Select optimal classification model;
The disaggregated model for determining user's selection, using the disaggregated model as object-class model.
Disaggregated model in the program includes categorised decision tree-model, Random Forest model, neural network model, Bayes
Classifier and SVM (Support Vector Machine, support vector machines) classifier etc., regression model include returning to determine
Plan tree-model, linear regression model (LRM) etc..
Further, the modeling procedure in the program further includes visualizing step, and the model construction program may be used also
It is executed by the processor, after the model parameter of each disaggregated model of the determination the step of, also realization following steps:
When receiving user based on the visual presentation instruction for visualizing the corresponding preset control creation of step, really
Determine the visual presentation mode of user setting;
The correlation for carrying out variable distribution to current data column is analyzed, and by analysis result according to user setting
Visual presentation mode is shown.Wherein, visual presentation mode includes box figure, pie chart, line chart, histogram, Q-Q figure (Q generation
Table quantile) etc., the type that user can according to need the data of displaying selects corresponding visual presentation mode.
Further, in one embodiment, a data file declustering can also be by the program using clustering algorithm
Multiple data subfiles.Then it is directed to different regression model or disaggregated model, multiple modeling procedures are constructed, by data Ziwen
Part is trained model respectively as the input data of multiple modeling procedures.Wherein, the classification in above-mentioned fractionation data file
Target data column in algorithm, with regression model or disaggregated model are not identical.User terminal transmission is received for example, executing and working as
Modeling instruction when, the step of showing pre-set control selection interface, is to transferring the function control selected with user from database
The corresponding code of part, and the step of being sequentially generated modeling procedure according to the execution between the functionality controls construct file declustering
The data of input are split as multiple data subfiles by sorting algorithm by process, will split obtained multiple data respectively
The step of file is inputted as data, executes this programme again respectively.
The model construction device that the present embodiment proposes is shown preparatory when receiving the modeling instruction of user terminal transmission
The control selection interface of setting, wherein include the control selection region of multiple modeling procedures, Duo Gejian in control selection interface
Mould step is for creating disaggregated model or regression model;Determine user from the control selection region in control selection interface be it is each
The functionality controls of modeling procedure selection, and sequence is executed between multiple functionality controls of determining user setting, wherein one is built
Mould step corresponds to one or more functionality controls;Code corresponding with functionality controls is transferred from database, and according to function
Execution between control is sequentially generated modeling procedure;When receive based in modeling procedure functionality controls triggering operational order
When, the corresponding code of the functionality controls is executed, until completing the training of disaggregated model or regression model, and determines model parameter.
The program pre-sets each functionality controls that modeling procedure may be used, and user can according to need the arbitrary function of selection
Control creates modeling procedure, even if, if user needs to change modeling procedure, only needing replacement function control after model construction completion
Part improves modeling efficiency without exploitation code again.
Optionally, in other examples, model construction program can also be divided into one or more module, and one
A or multiple modules are stored in memory 11, and are held by one or more processors (the present embodiment is by processor 12)
For row to complete the present invention, the so-called module of the present invention is the series of computation machine program instruction section for referring to complete specific function,
For implementation procedure of the descriptive model construction procedures in model construction device.
It is the program mould of the model construction program in one embodiment of model construction device of the present invention for example, referring to shown in Fig. 3
Block schematic diagram, in the embodiment, model construction program can be divided into showing interface module 10, control selecting module 20, stream
Journey determining module 30 and model construction module 40, illustratively:
Showing interface module 10 is used for: when receiving the modeling instruction of user terminal transmission, showing pre-set control
Part selection interface, wherein it include the control selection region of multiple modeling procedures in the control selection interface, it is the multiple to build
Mould step is for creating disaggregated model or regression model;
Control selecting module 20 is used for: determining that user builds from the control selection region in the control selection interface to be each
The functionality controls of mould step selection, and sequence is executed between the multiple functionality controls of determining user setting, wherein one
Modeling procedure corresponds to one or more functionality controls;
Process determining module 30 is used for: transfer the corresponding code of functionality controls selected with user from database, and according to
Execution between the functionality controls is sequentially generated modeling procedure;
Model construction module 40 is used for: when receive based in the modeling procedure functionality controls triggering operational order
When, the corresponding code of the functionality controls is executed, until completing the training of disaggregated model or regression model, and determines model parameter.
The journeys such as above-mentioned showing interface module 10, control selecting module 20, process determining module 30 and model construction module 40
Sequence module is performed realized functions or operations step and is substantially the same with above-described embodiment, and details are not described herein.
In addition, the embodiment of the present invention also proposes a kind of computer readable storage medium, the computer readable storage medium
On be stored with model construction program, the model construction program can be executed by one or more processors, to realize following operation:
When receiving the modeling instruction of user terminal transmission, pre-set control selection interface is shown, wherein described
It include the control selection region of multiple modeling procedures in control selection interface, the multiple modeling procedure is for creating classification mould
Type or regression model;
Determine user from the function control that the control selection region in the control selection interface is that each modeling procedure selects
Part, and sequence is executed between the multiple functionality controls of determining user setting, wherein a modeling procedure corresponds to one
Or multiple functionality controls;
Code corresponding with the functionality controls of user's selection is transferred from database, and according to holding between the functionality controls
Row is sequentially generated modeling procedure;
When receiving the operational order based on the functionality controls triggering in the modeling procedure, the functionality controls pair are executed
The code answered until completing the training of disaggregated model or regression model, and determines model parameter.The computer-readable storage of the present invention
Medium specific embodiment and above-mentioned each embodiment of model construction device and method are essentially identical, do not make tired state herein.
It should be noted that the serial number of the above embodiments of the invention is only for description, do not represent the advantages or disadvantages of the embodiments.And
The terms "include", "comprise" herein or any other variant thereof is intended to cover non-exclusive inclusion, so that packet
Process, device, article or the method for including a series of elements not only include those elements, but also including being not explicitly listed
Other element, or further include for this process, device, article or the intrinsic element of method.Do not limiting more
In the case where, the element that is limited by sentence "including a ...", it is not excluded that including process, device, the article of the element
Or there is also other identical elements in method.
Through the above description of the embodiments, those skilled in the art can be understood that above-described embodiment side
Method can be realized by means of software and necessary general hardware platform, naturally it is also possible to by hardware, but in many cases
The former is more preferably embodiment.Based on this understanding, technical solution of the present invention substantially in other words does the prior art
The part contributed out can be embodied in the form of software products, which is stored in one as described above
In storage medium (such as ROM/RAM, magnetic disk, CD), including some instructions are used so that terminal device (it can be mobile phone,
Computer, server or network equipment etc.) execute method described in each embodiment of the present invention.
The above is only a preferred embodiment of the present invention, is not intended to limit the scope of the invention, all to utilize this hair
Equivalent structure or equivalent flow shift made by bright specification and accompanying drawing content is applied directly or indirectly in other relevant skills
Art field, is included within the scope of the present invention.
Claims (10)
1. a kind of model building method, which is characterized in that the described method includes:
When receiving the modeling instruction of user terminal transmission, pre-set control selection interface is shown, wherein the control
Include the control selection region of multiple modeling procedures in selection interface, the multiple modeling procedure for create disaggregated model or
Regression model;
Determine the functionality controls that user selects from the control selection region in the control selection interface for each modeling procedure, and
Determine and execute sequence between the multiple functionality controls of user setting, wherein modeling procedure correspond to one or
Multiple functionality controls;
Code corresponding with the functionality controls of user's selection is transferred from database, and suitable according to the execution between the functionality controls
Sequence generates modeling procedure;
When receiving the operational order based on the functionality controls triggering in the modeling procedure, it is corresponding to execute the functionality controls
Code until completing the training of disaggregated model or regression model, and determines model parameter.
2. model building method as described in claim 1, which is characterized in that the modeling procedure successively includes data load step
Suddenly, data initialization step, data prediction step, characteristic processing step and model training step;Described ought receive is based on
When the operational order of the default step triggering in the modeling procedure, the default corresponding code of step is executed, until completing to divide
The training of class model or regression model, and the step of determining model parameter includes:
When receiving the data load instruction that user is triggered based on the functionality controls of data load step, obtains user and be based on number
According to the data file that load step uploads, loads and show the data file;
After receiving the data initialization instruction that user is triggered based on the functionality controls of data initialization step, the number is parsed
According to file, and determine the input feature vector column in the data file and target signature column;
When receiving the data prediction instruction that user is triggered based on the functionality controls of data prediction step, according to preset
Preprocessing algorithms pre-process the feature in input feature vector column and target signature column;
When receiving the characteristic processing instruction that user is triggered based on the functionality controls of characteristic processing step, according to pretreated
The weight of the corresponding feature of each characteristic series of input feature vector column count, and show the weight, so that user is according to the power of displaying
Feature is filtered again, and determines the input feature vector column of user's selection;
When receiving the model training instruction that user is triggered based on the functionality controls of model training step, user's selection is determined
Disaggregated model or regression model, the input feature vector column selected using user and the target signature arrange the training disaggregated model or
Regression model is to determine model parameter.
3. model building method as claimed in claim 2, which is characterized in that the parsing data file, and determine institute
It states the input feature vector in data file and arranges and include: with the step of target signature column
Identify the data type of the characteristic series and characteristic series in the data file, and based on the characteristic series and characteristic series recognized
Data type display data role's set interface;
Determine user's input feature vector column that role's set interface is selected from data column based on the data and target signature
Column.
4. model building method as claimed in claim 2 or claim 3, which is characterized in that if user selection disaggregated model have it is multiple,
The then disaggregated model of determining user's selection, the input feature vector column selected using user and the target signature are arranged described in training
The step of disaggregated model or regression model are to determine model parameter include:
It determines multiple disaggregated models of user's selection, is arranged and target signature column instruction using the input feature vector that user selects respectively
Practice the multiple disaggregated model, with the model parameter of each disaggregated model of determination;
After the step of model parameter of each disaggregated model of determination, the method also includes steps:
The accuracy rate and classification report of each disaggregated model are shown, so that user selects according to the accuracy rate and classification report of displaying
Optimal classification model;
The disaggregated model for determining user's selection, using the disaggregated model as object-class model.
5. model building method as claimed in claim 2 or claim 3, which is characterized in that the modeling procedure further includes visualization exhibition
Show step, the parsing data file, and determines the step of the input feature vector column in the data file and target signature column
After rapid, the method also includes steps:
When receiving user based on the visual presentation instruction for visualizing the corresponding functionality controls creation of step, determines and use
The visual presentation mode of family setting;
The correlation for carrying out variable distribution to current data column is analyzed, and by analysis result according to the visual of user setting
Change exhibition method to show.
6. a kind of model construction device, which is characterized in that described device includes memory and processor, is stored on the memory
There is the model construction program that can be run on the processor, is realized such as when the model construction program is executed by the processor
Lower step:
When receiving the modeling instruction of user terminal transmission, pre-set control selection interface is shown, wherein the control
Include the control selection region of multiple modeling procedures in selection interface, the multiple modeling procedure for create disaggregated model or
Regression model;
Determine the functionality controls that user selects from the control selection region in the control selection interface for each modeling procedure, and
Determine and execute sequence between the multiple functionality controls of user setting, wherein modeling procedure correspond to one or
Multiple functionality controls;
Code corresponding with the functionality controls of user's selection is transferred from database, and suitable according to the execution between the functionality controls
Sequence generates modeling procedure;
When receiving the operational order based on the functionality controls triggering in the modeling procedure, it is corresponding to execute the functionality controls
Code until completing the training of disaggregated model or regression model, and determines model parameter.
7. model construction device as claimed in claim 6, which is characterized in that the modeling procedure successively includes data load step
Suddenly, data initialization step, data prediction step, characteristic processing step and model training step;Described ought receive is based on
When the operational order of the default step triggering in the modeling procedure, the default corresponding code of step is executed, until completing to divide
The training of class model or regression model, and the step of determining model parameter includes:
When receiving the data load instruction that user is triggered based on the functionality controls of data load step, obtains user and be based on number
According to the data file that load step uploads, loads and show the data file;
After receiving the data initialization instruction that user is triggered based on the functionality controls of data initialization step, the number is parsed
According to file, and determine the input feature vector column in the data file and target signature column;
When receiving the data prediction instruction that user is triggered based on the functionality controls of data prediction step, according to preset
Preprocessing algorithms pre-process the feature in input feature vector column and target signature column;
When receiving the characteristic processing instruction that user is triggered based on the functionality controls of characteristic processing step, according to pretreated
The weight of the corresponding feature of each characteristic series of input feature vector column count, and show the weight, so that user is according to the power of displaying
Feature is filtered again, and determines the input feature vector column of user's selection;
When receiving the model training instruction that user is triggered based on the functionality controls of model training step, user's selection is determined
Disaggregated model or regression model, the input feature vector column selected using user and the target signature arrange the training disaggregated model or
Regression model is to determine model parameter.
8. model construction device as claimed in claim 7, which is characterized in that the parsing data file, and determine institute
It states the input feature vector in data file and arranges and include: with the step of target signature column
Identify the data type of the characteristic series and characteristic series in the data file, and based on the characteristic series and characteristic series recognized
Data type display data role's set interface;
Determine user's input feature vector column that role's set interface is selected from data column based on the data and target signature
Column.
9. model construction device as claimed in claim 7 or 8, which is characterized in that if user selection disaggregated model have it is multiple,
The then disaggregated model of determining user's selection, the input feature vector column selected using user and the target signature are arranged described in training
The step of disaggregated model or regression model are to determine model parameter include:
It determines multiple disaggregated models of user's selection, is arranged and target signature column instruction using the input feature vector that user selects respectively
Practice the multiple disaggregated model, with the model parameter of each disaggregated model of determination;
The model construction program can also be executed by the processor, in the model parameter of each disaggregated model of the determination
After step, following steps are also realized:
The accuracy rate and classification report of each disaggregated model are shown, so that user selects according to the accuracy rate and classification report of displaying
Optimal classification model;
The disaggregated model for determining user's selection, using the disaggregated model as object-class model.
10. a kind of computer readable storage medium, which is characterized in that be stored with model structure on the computer readable storage medium
Program is built, the model construction program can be executed by one or more processor, to realize as any in claim 1 to 5
The step of model building method described in item.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811524934.XA CN109783859A (en) | 2018-12-13 | 2018-12-13 | Model building method, device and computer readable storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811524934.XA CN109783859A (en) | 2018-12-13 | 2018-12-13 | Model building method, device and computer readable storage medium |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109783859A true CN109783859A (en) | 2019-05-21 |
Family
ID=66496873
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811524934.XA Pending CN109783859A (en) | 2018-12-13 | 2018-12-13 | Model building method, device and computer readable storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109783859A (en) |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110750909A (en) * | 2019-10-25 | 2020-02-04 | 中国建设银行股份有限公司 | Modeling method, device, equipment and medium based on model building platform |
CN110868471A (en) * | 2019-11-18 | 2020-03-06 | 北京小米移动软件有限公司 | Novel equipment construction method and device and storage medium |
CN111241746A (en) * | 2020-01-09 | 2020-06-05 | 深圳前海微众银行股份有限公司 | Forward model selection method, apparatus and readable storage medium |
CN111291071A (en) * | 2020-01-21 | 2020-06-16 | 北京字节跳动网络技术有限公司 | Data processing method and device and electronic equipment |
CN111507066A (en) * | 2020-04-21 | 2020-08-07 | 北京思特奇信息技术股份有限公司 | Method and system for uploading, analyzing and warehousing loaded files |
CN111565487A (en) * | 2020-04-21 | 2020-08-21 | 深圳海令科技有限公司 | PLC-BUS-based ten-thousand-stage light-operated adjusting system |
CN112445462A (en) * | 2020-11-16 | 2021-03-05 | 北京思特奇信息技术股份有限公司 | Artificial intelligence modeling platform and method based on object-oriented design |
CN113050938A (en) * | 2021-03-08 | 2021-06-29 | 杭州海康机器人技术有限公司 | Visual software development system, method, device and computer storage medium |
CN114356295A (en) * | 2021-12-24 | 2022-04-15 | 北京有竹居网络技术有限公司 | Three-dimensional model construction method and device and electronic equipment |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106779087A (en) * | 2016-11-30 | 2017-05-31 | 福建亿榕信息技术有限公司 | A kind of general-purpose machinery learning data analysis platform |
CN107103050A (en) * | 2017-03-31 | 2017-08-29 | 海通安恒(大连)大数据科技有限公司 | A kind of big data Modeling Platform and method |
CN107169575A (en) * | 2017-06-27 | 2017-09-15 | 北京天机数测数据科技有限公司 | A kind of modeling and method for visualizing machine learning training pattern |
CN107633254A (en) * | 2017-07-25 | 2018-01-26 | 平安科技(深圳)有限公司 | Establish device, method and the computer-readable recording medium of forecast model |
CN108416825A (en) * | 2017-07-25 | 2018-08-17 | 平安科技(深圳)有限公司 | Generating means, method and the computer readable storage medium of Dynamic Graph |
CN108446225A (en) * | 2018-03-06 | 2018-08-24 | 平安科技(深圳)有限公司 | Creating device, method and the computer readable storage medium of analog interface |
CN108459792A (en) * | 2018-01-25 | 2018-08-28 | 阿里巴巴集团控股有限公司 | A kind of flow switching method, device and computer equipment |
-
2018
- 2018-12-13 CN CN201811524934.XA patent/CN109783859A/en active Pending
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106779087A (en) * | 2016-11-30 | 2017-05-31 | 福建亿榕信息技术有限公司 | A kind of general-purpose machinery learning data analysis platform |
CN107103050A (en) * | 2017-03-31 | 2017-08-29 | 海通安恒(大连)大数据科技有限公司 | A kind of big data Modeling Platform and method |
CN107169575A (en) * | 2017-06-27 | 2017-09-15 | 北京天机数测数据科技有限公司 | A kind of modeling and method for visualizing machine learning training pattern |
CN107633254A (en) * | 2017-07-25 | 2018-01-26 | 平安科技(深圳)有限公司 | Establish device, method and the computer-readable recording medium of forecast model |
CN108416825A (en) * | 2017-07-25 | 2018-08-17 | 平安科技(深圳)有限公司 | Generating means, method and the computer readable storage medium of Dynamic Graph |
CN108459792A (en) * | 2018-01-25 | 2018-08-28 | 阿里巴巴集团控股有限公司 | A kind of flow switching method, device and computer equipment |
CN108446225A (en) * | 2018-03-06 | 2018-08-24 | 平安科技(深圳)有限公司 | Creating device, method and the computer readable storage medium of analog interface |
Cited By (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110750909A (en) * | 2019-10-25 | 2020-02-04 | 中国建设银行股份有限公司 | Modeling method, device, equipment and medium based on model building platform |
CN110750909B (en) * | 2019-10-25 | 2023-10-27 | 中国建设银行股份有限公司 | Modeling method, device, equipment and medium based on model construction platform |
US11729305B2 (en) | 2019-11-18 | 2023-08-15 | Beijing Xiaomi Mobile Software Co., Ltd. | Method and apparatus for constructing novel device |
CN110868471A (en) * | 2019-11-18 | 2020-03-06 | 北京小米移动软件有限公司 | Novel equipment construction method and device and storage medium |
CN111241746A (en) * | 2020-01-09 | 2020-06-05 | 深圳前海微众银行股份有限公司 | Forward model selection method, apparatus and readable storage medium |
CN111241746B (en) * | 2020-01-09 | 2024-01-26 | 深圳前海微众银行股份有限公司 | Forward model selection method, apparatus, and readable storage medium |
CN111291071A (en) * | 2020-01-21 | 2020-06-16 | 北京字节跳动网络技术有限公司 | Data processing method and device and electronic equipment |
CN111291071B (en) * | 2020-01-21 | 2023-10-17 | 北京字节跳动网络技术有限公司 | Data processing method and device and electronic equipment |
CN111565487A (en) * | 2020-04-21 | 2020-08-21 | 深圳海令科技有限公司 | PLC-BUS-based ten-thousand-stage light-operated adjusting system |
CN111507066B (en) * | 2020-04-21 | 2024-01-12 | 北京思特奇信息技术股份有限公司 | Method and system for uploading and loading files and analyzing and warehousing |
CN111507066A (en) * | 2020-04-21 | 2020-08-07 | 北京思特奇信息技术股份有限公司 | Method and system for uploading, analyzing and warehousing loaded files |
CN112445462A (en) * | 2020-11-16 | 2021-03-05 | 北京思特奇信息技术股份有限公司 | Artificial intelligence modeling platform and method based on object-oriented design |
CN113050938A (en) * | 2021-03-08 | 2021-06-29 | 杭州海康机器人技术有限公司 | Visual software development system, method, device and computer storage medium |
CN114356295A (en) * | 2021-12-24 | 2022-04-15 | 北京有竹居网络技术有限公司 | Three-dimensional model construction method and device and electronic equipment |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109783859A (en) | Model building method, device and computer readable storage medium | |
CN109165249B (en) | Data processing model construction method and device, server and user side | |
CN108427698A (en) | Updating device, method and the computer readable storage medium of prediction model | |
CN107633254A (en) | Establish device, method and the computer-readable recording medium of forecast model | |
CN108090516A (en) | Automatically generate the method and system of the feature of machine learning sample | |
CN107679455A (en) | Target tracker, method and computer-readable recording medium | |
CN110458360B (en) | Method, device, equipment and storage medium for predicting hot resources | |
US20170300598A1 (en) | System and method for designing a product and manufacturing a product | |
CN107004185A (en) | The pipeline generation of the control actuated for data flow | |
CN109325053A (en) | Data processing method, device and the computer readable storage medium of reporting system | |
CN109074368A (en) | Use the graphics frame of isomery social networks | |
CN104823187A (en) | Displaying sort results on mobile computing device | |
US20200122039A1 (en) | Method and system for a behavior generator using deep learning and an auto planner | |
US20230274480A1 (en) | Method and system for populating a digital environment using a semantic map | |
CN107818492A (en) | Products Show device, method and computer-readable recording medium | |
KR20230014802A (en) | Method for recommending designated items | |
CN115022098B (en) | Artificial intelligence safety target range content recommendation method, device and storage medium | |
CN109522923A (en) | Customer address polymerization, device and computer readable storage medium | |
CN109934194A (en) | Picture classification method, edge device, system and storage medium | |
CN103294805A (en) | Creation method and device for data warehouse personalized dimension table | |
CN109598285A (en) | A kind of processing method of model, device and equipment | |
CN109871205A (en) | GUI code method of adjustment, device, computer installation and storage medium | |
CN108427586B (en) | Using push terminal, method and the computer readable storage medium of theme | |
US20230326206A1 (en) | Visual tag emerging pattern detection | |
KR20210133090A (en) | Electronic device for providing information realted to failure of product and method for operating thereof |
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 |