CN108038012A - Data calibration method and device, electronic equipment and computer-readable recording medium - Google Patents
Data calibration method and device, electronic equipment and computer-readable recording medium Download PDFInfo
- Publication number
- CN108038012A CN108038012A CN201711288199.2A CN201711288199A CN108038012A CN 108038012 A CN108038012 A CN 108038012A CN 201711288199 A CN201711288199 A CN 201711288199A CN 108038012 A CN108038012 A CN 108038012A
- Authority
- CN
- China
- Prior art keywords
- data
- operational model
- operation result
- normalized
- reception
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/006—Identification
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Quality & Reliability (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Stored Programmes (AREA)
Abstract
The present invention provides a kind of data calibration method and device, electronic equipment and computer-readable recording medium.The data calibration method can realize the transmission and reception of data between different running environment by data-interface, to realize the calibration to the data of nonidentity operation model.Versatility of the present invention is good, and conveniently transplanting, convenient extension, convenient fractionation, is adapted to different running environment.
Description
Technical field
The present invention relates to technical field of data processing, more particularly to a kind of data calibration method and device, electronic equipment and
Computer-readable recording medium.
Background technology
With memory technology, Internet technology and its rapid development of application, data become increasing, more and more multiple
Miscellaneous, the species of data is also more and more, and therefore, data calibration also just becomes more and more difficult therewith.
In prior art, the method for data calibration is typically to be required for specific, design temporarily a kind of method or
Device is realized, lacks versatility.Moreover, these method or apparatus also it is less consider data calibration efficiency.Big data background
Under data calibration, can be through all too many levels of development and application, versatility and efficiency are crucial.
The content of the invention
In view of the foregoing, it is necessary to a kind of data calibration method and device, electronic equipment are provided and computer-readable deposited
Storage media, not only versatility is good, and conveniently transplanting, convenient extension, convenient fractionation, is adapted to different running environment, and in fact
Now to the data calibration of nonidentity operation model.
A kind of data calibration method, the described method includes：
Pending data is selected from the data set of reception；
The pending data is converted into the data of at least one data format, at least one data format includes
The input data form of first operational model and the input data form of the second operational model, first operational model with it is described
Second operational model is used for realization identical algorithm；
The first input data input of the input data form of first operational model will be met in transformed data
Computing is carried out into first operational model, obtains the first operation result；
Data-interface is called, the input data form of second operational model will be met in the transformed data
Second input data is inputted carries out computing into second operational model, obtains the second operation result；
Obtain first operation result and second operation result；
Second operation result and first operation result are normalized；
According to first operation result after normalized and second operation result after normalized, really
The calibration result of fixed second operational model.
Preferred embodiment according to the present invention, selection pending data includes in the data set from reception：
Total data in the data set of the reception is determined as the pending data；Or
Data in the data set of the reception are ranked up according to the size of data, the data of presetting digit capacity will be come
It is determined as the pending data；Or
The data of preset kind in the data set of the reception are determined as the pending data；Or
The data in the data set of the reception with default mark are determined as the pending data.
Preferred embodiment according to the present invention, first operational model are to realize the algorithm using the first programming language,
Second operational model is to realize the algorithm, first programming language and the described second programming using the second programming language
Language is different.
Preferred embodiment according to the present invention, before calling data-interface, the method further includes：
By calling dynamic link library to form the data-interface.
Preferred embodiment according to the present invention, second operational model, which has, accelerates module, and the acceleration module is used to add
The speed of fast computing.
Preferred embodiment according to the present invention, at first operation result and normalization according to after normalized
Second operation result after reason, determining the calibration result of second operational model includes：
First operation result after second operation result after normalized and normalized is carried out
Contrast, obtains calculation error；
When the calculation error is in default error range, determine that second operational model is correct；Or
When the calculation error is not in the default error range, determine that second operational model is inaccurate.
Preferred embodiment according to the present invention, at first operation result and normalization according to after normalized
Second operation result after reason, determines that the calibration result of second operational model further includes：
Using default error analysis method by after second operation result after normalized and normalized
First operation result is contrasted, and obtains the error curve of second operation result；
Analyze the convergence of the error curve；
When the error curve converges on preset value, determine that the second operational model computing is correct；Or
When the error curve does not converge on the preset value, determine that the second operational model computing is inaccurate.
A kind of data calibration device, described device include：
Selecting unit, for selecting pending data from the data set of reception；
Converting unit, for the pending data to be converted into the data of at least one data format, described at least one
The input data form of input data form and second operational model of the kind data format including the first operational model, described first
Operational model is used for realization identical algorithm with second operational model；
Arithmetic element, for the first of the input data form of first operational model will to be met in transformed data
Input data is inputted carries out computing into first operational model, obtains the first operation result；
Call unit, for calling data-interface, will meet second operational model in the transformed data
Second input data of input data form is inputted carries out computing into second operational model, obtains the second operation result；
Acquiring unit, for obtaining first operation result and second operation result；
Processing unit, for second operation result and first operation result to be normalized；
Determination unit, for according to described after first operation result after normalized and normalized
Two operation results, determine the calibration result of second operational model.
Preferred embodiment according to the present invention, the selecting unit are specifically used for：
Total data in the data set of the reception is determined as the pending data；Or
Data in the data set of the reception are ranked up according to the size of data, the data of presetting digit capacity will be come
It is determined as the pending data；Or
The data of preset kind in the data set of the reception are determined as the pending data；Or
The data in the data set of the reception with default mark are determined as the pending data.
Preferred embodiment according to the present invention, first operational model are to realize the algorithm using the first programming language,
Second operational model is to realize the algorithm, first programming language and the described second programming using the second programming language
Language is different.
Preferred embodiment according to the present invention, described device further include：
Data interface unit, for before data-interface is called, by calling dynamic link library to form the data-interface.
Preferred embodiment according to the present invention, second operational model, which has, accelerates module, and the acceleration module is used to add
The speed of fast computing.
Preferred embodiment according to the present invention, the determination unit are specifically used for：
First operation result after second operation result after normalized and normalized is carried out
Contrast, obtains calculation error；
When the calculation error is in default error range, determine that second operational model is correct；Or
When the calculation error is not in the default error range, determine that second operational model is inaccurate.
Preferred embodiment according to the present invention, the determination unit are specifically additionally operable to：
Using default error analysis method by after second operation result after normalized and normalized
First operation result is contrasted, and obtains the error curve of second operation result；
Analyze the convergence of the error curve；
When the error curve converges on preset value, determine that the second operational model computing is correct；Or
When the error curve does not converge on the preset value, determine that the second operational model computing is inaccurate.
A kind of electronic equipment, the electronic equipment include：
Memory, stores at least one instruction；And
Processor, performs the instruction that is stored in the memory to realize the data calibration method.
A kind of computer-readable recording medium, is stored with least one instruction, institute in the computer-readable recording medium
At least one instruction is stated to be performed by the processor in electronic equipment to realize the data calibration method.
As can be seen from the above technical solutions, the present invention can select pending data from the data set of reception；By described in
Pending data is converted into the data of at least one data format, and at least one data format includes the first operational model
The input data form of input data form and the second operational model, first operational model are used with second operational model
In realizing identical algorithm；The first input of the input data form of first operational model will be met in transformed data
Data are inputted carries out computing into first operational model, obtains the first operation result；Data-interface is called, by the conversion
The second input data for meeting the input data form of second operational model in data afterwards is inputted to second computing
Computing is carried out in model, obtains the second operation result；Obtain first operation result and second operation result；To described
Second operation result and first operation result are normalized；According to the first computing knot after normalized
Second operation result after fruit and normalized, determines the calibration result of second operational model.Therefore, it is of the invention
Versatility is good, and conveniently transplanting, convenient extension, convenient fractionation, is adapted to different running environment, and realize to nonidentity operation
The data calibration of model.
Brief description of the drawings
Fig. 1 is the flow chart of the preferred embodiment of data calibration method of the present invention.
Fig. 2 is the functional block diagram of the preferred embodiment of data calibration device of the present invention.
Fig. 3 is the structure diagram of the electronic equipment for the preferred embodiment that the present invention realizes data calibration method.
Embodiment
In order to make the object, technical solutions and advantages of the present invention clearer, below in conjunction with the accompanying drawings with specific embodiment pair
The present invention is described in detail.
As shown in Figure 1, it is the flow chart of the preferred embodiment of data calibration method of the present invention., should according to different demands
The order of step can change in flow chart, and some steps can be omitted.
The data calibration method is applied in one or more electronic equipment, and the electronic equipment is that one kind can be by
According to the instruction for being previously set or storing, the equipment of automatic progress numerical computations and/or information processing, its hardware includes but not limited to
Microprocessor, application-specific integrated circuit (Application Specific Integrated Circuit, ASIC), programmable gate
Array (Field-Programmable Gate Array, FPGA), digital processing unit (Digital Signal
Processor, DSP), embedded device etc..
The electronic equipment can be the electronic product that any type can carry out human-computer interaction with user, for example, personal meter
Calculation machine, tablet computer, smart mobile phone, personal digital assistant (Personal Digital Assistant, PDA), game machine, friendship
Mutual formula Web TV (Internet Protocol Television, IPTV), intellectual Wearable etc..
The electronic equipment can also include the network equipment and/or user equipment.Wherein, the network equipment includes, but
It is not limited to single network server, the server group of multiple webservers composition or based on cloud computing (Cloud
Computing the cloud being made of a large amount of hosts or the webserver).
Network residing for the electronic equipment include but not limited to internet, wide area network, Metropolitan Area Network (MAN), LAN, it is virtual specially
With network (Virtual Private Network, VPN) etc..
S10, the electronic equipment select pending data from the data set of reception.
In at least one embodiment of the present invention, pending number is selected in data set of the electronic equipment from reception
According to before, the method further includes：
The electronic equipment receives data set.
In at least one embodiment of the present invention, the data set includes, but are not limited to：Communicated with the electronic equipment
Data in the database of letter, the parameter to match with the data in the database, the mark of data in the database
Deng.
In at least one embodiment of the present invention, the electronic equipment selects pending data from the data set of reception
Mode include, but are not limited to following one or more kinds of combination：
(1) total data in the data set of the reception is determined as the pending data by the electronic equipment.
In at least one embodiment of the present invention, when the data volume of the data set of the reception meets process demand,
Then the total data in the data set of the reception is determined as the pending data by the electronic equipment.
(2) data in the data set of the reception are ranked up by the electronic equipment according to the size of data, will be arranged
It is determined as the pending data in the data of presetting digit capacity.
In at least one embodiment of the present invention, the electronic equipment can be by the part in the data set of the reception
Data are as the pending data.
Preferably, the electronic equipment can arrange the data in the data set of the reception according to the size of data
Sequence, and the data of presetting digit capacity will be come as the pending data.
In at least one embodiment of the present invention, the presetting digit capacity includes, but are not limited to：First 50, latter 50 etc.
Deng, the electronic equipment can make choice according to process demand, the present invention be not restricted.
In this way, in order to save the time of data processing, or in order to adapt to the data-handling capacity of nonidentity operation model,
The electronic equipment can be according to the size of data using the partial data in the data set of the reception as the pending number
According to making the selection of the pending data more flexibly and directly perceived.
In addition, when the data in the data set of the reception are ranked up by the electronic equipment according to the size of data,
Can be according to the order of data from small to large, can also be according to the order of data from big to small, the present invention is not restricted.It can manage
Solution, according to the difference of sortord, the presetting digit capacity can also be set accordingly.
(3) data of preset kind in the data set of the reception are determined as the pending number by the electronic equipment
According to.
In at least one embodiment of the present invention, the electronic equipment by the data in the data set of the reception according to
Type is classified, and the data of the preset kind are determined as the pending data by the electronic equipment.
Such as：When the data in the data set of the reception include picture, video when different type, at this time, it is not
All types of data are required for being handled, if the data for being actually needed processing are the data of picture/mb-type, the electricity
Sub- equipment can select the data of picture/mb-type as the pending number according to being actually needed from the data set of the reception
According to so as to avoid the unnecessary loss of time, making the more efficient of data processing.
Certainly, can also be selected at the same time from the data set of the reception according to different process demands, the electronic equipment
The data of picture/mb-type and the data of video type are selected not to be restricted as the pending data, the present invention.
(4) data in the data set of the reception with default mark are determined as described pending by the electronic equipment
Data.
In at least one embodiment of the present invention, the electronic equipment can be by the data in the data set of the reception
It is labeled.Such as：In a pictures, the electronic equipment can mark out specifying in the picture in the form of square frame
The face of personage, either, the electronic equipment can mark out predeterminated position in the picture etc. in the form of circular frame.
In at least one embodiment of the present invention, the electronic equipment can select tool from the data set of the reception
There are the data of identical mark as the pending data, such as：The electronic equipment can be from the data set of the reception
Selection is using the picture of circular frame mark face as the pending data.In this way, the processing procedure of data is then more added with pin
To property, accuracy is not only improved, while saved time cost.
The pending data is converted into the data of at least one data format by S11, the electronic equipment, it is described at least
A kind of input data form of data format including the first operational model and the input data form of the second operational model, described the
One operational model is used for realization identical algorithm with second operational model.
In at least one embodiment of the present invention, since the data format of data handled by different operational models is general
Difference, therefore, in order to make different operational models the pending data can be used to carry out computing, the electronic equipment can
The pending data is converted into by least one data lattice with the data format of the data according to handled by different operational models
The data of formula, so that the technical program has more preferable versatility.
In at least one embodiment of the present invention, at least one data format includes, but are not limited to：Described first
Input data form of the input data form of operational model and second operational model etc..Wherein, the first computing mould
Type is used for realization identical algorithm with second operational model.
Specifically, first operational model is to realize the algorithm, the second computing mould using the first programming language
Type is to realize the algorithm using the second programming language, and first programming language is different from second programming language.
In at least one embodiment of the present invention, first operational model is a kind of mould realized with software language
Type, specifically, first programming language include, but are not limited to：C language, C++, MATLAB etc., the present invention do not limit herein
System.In addition, learnt according to the lot of experiment validation of related technical personnel, the computing obtained by the first operational model computing
The result is that standard as a result, can be as the basis of data calibration.
In at least one embodiment of the present invention, second operational model is a kind of mould realized with hardware language
Type, specifically, second programming language include, but are not limited to：The programming language of embedded system, the programming language of FPGA,
Programming language of CPLD etc., the present invention are not restricted.
In at least one embodiment of the present invention, first operational model is used for realization with second operational model
Identical algorithm, in this way, by the way that the operation result of second operational model and the first operational model computing obtained
Standard results are compared, it is possible to determine whether the computing of second operational model is accurate.
S12, the electronic equipment will meet the input data form of first operational model in transformed data
First input data is inputted carries out computing into first operational model, obtains the first operation result.
In at least one embodiment of the present invention, the conversion of the data format of the pending data can use and institute
State the identical programming language of the first operational model to realize, in this way, the electronic equipment will directly can accord with transformed data
The first input data for closing the input data form of first operational model is inputted into first operational model and transported
Calculate, obtain the first operation result.
In other embodiments, the electronic equipment can also use the programming language different from first operational model
Realize the conversion of the data format of the pending data, in this way, the electronic equipment will meet in transformed data it is described
, can when first input data of the input data form of the first operational model inputs into first operational model progress computing
The input of first input data is realized in a manner of by calling corresponding data-interface.
S13, the electronic equipment call data-interface, second operational model will be met in the transformed data
The second input data of input data form input and carry out computing into second operational model, obtain the second computing knot
Fruit.
In at least one embodiment of the present invention, will meet in the electronic equipment in the transformed data described
During second input data of the input data form of the second operational model is inputted into second operational model, due to
It is related to transmission of the data from software runtime environment to hardware running environment, therefore, the electronic equipment needs to call corresponding
Data-interface.Specifically, the electronic equipment is by calling dynamic link library to form the data-interface.The dynamic link library
The transmission and reception of data are used for realization, the electronic equipment calls the dynamic link library can be by the transformed data
In meet second operational model the second input data of input data form input into second operational model.
It should be noted that the present invention realizes that language is not restricted to the dynamic link library, such as：The electronics is set
It is standby the language such as C language, C++ to be used to write the dynamic link library.
In other embodiments, the data-interface can be used for data between the model that different software programming language is realized
Transmission, the present invention be not restricted.
In this way, by calling data-interface, make it is independent mutually between each operational model, to reach convenient transplanting, convenient expand
Exhibition, the convenient technique effect split, while different running environment is adapted to, and realize the data school to nonidentity operation model
It is accurate.
In at least one embodiment of the present invention, second operational model, which has, accelerates module, the acceleration module
For accelerating the speed of computing.
Specifically, it is described to accelerate the hardware that module be the acceleration data operation realized by hardware program language
Module, it is in the present embodiment, described to accelerate module to be used for the arithmetic speed for accelerating second operational model.
Specifically, it is described acceleration module can be with FPGA realize or a Software Development Kit, or
It is data operation model under a deep learning frame etc., this is not restricted by the present invention.
S14, the electronic equipment obtain first operation result and second operation result.
In at least one embodiment of the present invention, running environment is mentioned with reference to programming language, the electronic equipment can lead to
The acquisition for calling the mode of corresponding data-interface to realize first operation result and second operation result is crossed, the present invention
It is not restricted.
Second operation result and first operation result is normalized in S15, the electronic equipment.
In at least one embodiment of the present invention, in order to eliminate shadow of the different data formats to data results
Ring, second operation result and first operation result is normalized in the electronic equipment, so that described the
Two operation results and first operation result are comparable.
S16, the electronic equipment is according to first operation result after normalized and the institute after normalized
The second operation result is stated, determines the calibration result of second operational model.
In at least one embodiment of the present invention, the electronic equipment is according to first computing after normalized
As a result second operation result and after normalized, determining the calibration result of second operational model includes：
The electronic equipment is by described first after second operation result after normalized and normalized
Operation result is contrasted, and obtains calculation error.
Preferably, when the calculation error is in default error range, the electronic equipment determines second computing
Model is correct；When the calculation error is not in the default error range, the electronic equipment determines second computing
Model is inaccurate.
In at least one embodiment of the present invention, the default error range can be according to known in correlative technology field
Empirical value be determined.
In at least one embodiment of the present invention, the electronic equipment is according to first computing after normalized
As a result second operation result and after normalized, determines that the calibration result of second operational model further includes：
The electronic equipment is using default error analysis method by second operation result after normalized with returning
First operation result after one change processing is contrasted, and obtains the error curve of second operation result, the electronics
The convergence of error curve described in device analysis, and when the error curve converges on preset value, determine second fortune
It is correct to calculate model calculation；When the error curve does not converge on the preset value, the second operational model computing is determined not
Accurately.
In at least one embodiment of the present invention, the default error analysis method includes, but are not limited to：Mean square error
Analysis method, root-mean-square error analysis method, standard deviation analysis method etc., the present invention to the default error analysis method not
It is restricted.
In at least one embodiment of the present invention, the preset value be a less value (such as：1st, 0.89 etc.), this hair
It is bright not to be restricted.
In conclusion the present invention can select pending data from the data set of reception；The pending data is changed
Into the data of at least one data format, the input data form of at least one data format including the first operational model and
The input data form of second operational model, first operational model are used for realization identical calculation with second operational model
Method；First input data of the input data form for meeting first operational model in transformed data is inputted to described
Computing is carried out in first operational model, obtains the first operation result；Data-interface is called, will be met in the transformed data
Second input data of the input data form of second operational model is inputted carries out computing into second operational model,
Obtain the second operation result；Obtain first operation result and second operation result；To second operation result and
First operation result is normalized；According to first operation result and normalized after normalized
Second operation result afterwards, determines the calibration result of second operational model.Therefore, versatility of the present invention is good, and side
Just transplant, convenient extension, conveniently fractionation, be adapted to different running environment, and realize the data school to nonidentity operation model
It is accurate.
As shown in Fig. 2, it is the functional block diagram of the preferred embodiment of data calibration device of the present invention.The data calibration dress
Putting 11 includes selecting unit 110, converting unit 111, arithmetic element 112, call unit 113, acquiring unit 114, processing unit
115th, determination unit 116, data interface unit 117.Module/unit alleged by the present invention refers to that one kind can be by the institute of processor 13
Perform, and the series of computation machine program segment of fixed function can be completed, it is stored in memory 12.In the present embodiment
In, the function on each module/unit will be described in detail in follow-up embodiment.
Selecting unit 110 selects pending data from the data set of reception.
In at least one embodiment of the present invention, from selecting to treat in data set of the selecting unit 110 from reception
Before managing data, the method further includes：
The electronic equipment receives data set.
In at least one embodiment of the present invention, the data set includes, but are not limited to：Communicated with the electronic equipment
Data in the database of letter, the parameter to match with the data in the database, the mark of data in the database
Deng.
In at least one embodiment of the present invention, the selecting unit 110 selects pending from the data set of reception
The mode of data includes, but are not limited to following one or more kinds of combination：
(1) total data in the data set of the reception is determined as the pending data by the selecting unit 110.
In at least one embodiment of the present invention, when the data volume of the data set of the reception meets process demand,
Then the total data in the data set of the reception is determined as the pending data by the selecting unit 110.
(2) data in the data set of the reception are ranked up by the selecting unit 110 according to the size of data, will
The data for coming presetting digit capacity are determined as the pending data.
In at least one embodiment of the present invention, the selecting unit 110 can be by the data set of the reception
Partial data is as the pending data.
Preferably, the selecting unit 110 can to the data in the data set of the reception according to data size into
Row sequence, and the data of presetting digit capacity will be come as the pending data.
In at least one embodiment of the present invention, the presetting digit capacity includes, but are not limited to：First 50, latter 50 etc.
Deng, the selecting unit 110 can make choice according to process demand, the present invention be not restricted.
In this way, in order to save the time of data processing, or in order to adapt to the data-handling capacity of nonidentity operation model,
The selecting unit 110 can be according to the size of data using the partial data in the data set of the reception as described pending
Data, make the selection of the pending data more flexibly and directly perceived.
In addition, the data in the data set of the reception are ranked up by the selecting unit 110 according to the size of data
When, can be according to the order of data from small to large, can also be according to the order of data from big to small, the present invention is not restricted.Can
With understanding, according to the difference of sortord, the presetting digit capacity can also be set accordingly.
(3) data of preset kind in the data set of the reception are determined as described pending by the selecting unit 110
Data.
In at least one embodiment of the present invention, the selecting unit 110 is by the data in the data set of the reception
Classify according to type, the data of the preset kind are determined as the pending data by the selecting unit 110.
Such as：When the data in the data set of the reception include picture, video when different type, at this time, it is not
All types of data are required for being handled, if the data for being actually needed processing are the data of picture/mb-type, the choosing
The data of picture/mb-type can be selected as described pending according to being actually needed from the data set of the reception by selecting unit 110
Data, so as to avoid the unnecessary loss of time, make the more efficient of data processing.
Certainly, can also be from the data set of the reception at the same time according to different process demands, the selecting unit 110
The data of picture/mb-type and the data of video type are selected not to be restricted as the pending data, the present invention.
(4) data in the data set of the reception with default mark are determined as described treat by the selecting unit 110
Handle data.
In at least one embodiment of the present invention, the selecting unit 110 can be by the data set of the reception
Data are labeled.Such as：In a pictures, the electronic equipment can be marked out in the picture in the form of square frame
The face of designated person, either, the selecting unit 110 can be marked out default in the picture in the form of circular frame
Position etc..
In at least one embodiment of the present invention, the selecting unit 110 can be selected from the data set of the reception
The data with identical mark are selected as the pending data, such as：The selecting unit 110 can be from the number of the reception
The picture of face is marked as the pending data using circular frame according to selection is concentrated.In this way, the processing procedure of data is then more
Added with specific aim, accuracy is not only improved, while saved time cost.
The pending data is converted into the data of at least one data format, at least one by converting unit 111
Data format includes the input data form of the first operational model and the input data form of the second operational model, first fortune
Calculate model and be used for realization identical algorithm with second operational model.
In at least one embodiment of the present invention, since the data format of data handled by different operational models is general
Difference, therefore, in order to make different operational models the pending data can be used to carry out computing, the converting unit
111 can the data format of data according to handled by different operational models the pending data is converted at least one number
According to the data of form, so that the technical program has more preferable versatility.
In at least one embodiment of the present invention, at least one data format includes, but are not limited to：Described first
Input data form of the input data form of operational model and second operational model etc..Wherein, the first computing mould
Type is used for realization identical algorithm with second operational model.
Specifically, first operational model is to realize the algorithm, the second computing mould using the first programming language
Type is to realize the algorithm using the second programming language, and first programming language is different from second programming language.
In at least one embodiment of the present invention, first operational model is a kind of mould realized with software language
Type, specifically, first programming language include, but are not limited to：C language, C++, MATLAB etc., the present invention do not limit herein
System.In addition, learnt according to the lot of experiment validation of related technical personnel, the computing obtained by the first operational model computing
The result is that standard as a result, can be as the basis of data calibration.
In at least one embodiment of the present invention, second operational model is a kind of mould realized with hardware language
Type, specifically, second programming language include, but are not limited to：The programming language of embedded system, the programming language of FPGA,
Programming language of CPLD etc., the present invention are not restricted.
In at least one embodiment of the present invention, first operational model is used for realization with second operational model
Identical algorithm, in this way, by the way that the operation result of second operational model and the first operational model computing obtained
Standard results are compared, it is possible to determine whether the computing of second operational model is accurate.
It is first defeated will to meet the input data form of first operational model for arithmetic element 112 in transformed data
Enter data and input and carry out computing into first operational model, obtain the first operation result.
In at least one embodiment of the present invention, the conversion of the data format of the pending data can use and institute
State the identical programming language of the first operational model to realize, in this way, the arithmetic element 112 can be directly by transformed data
The first input data for meeting the input data form of first operational model is inputted into first operational model and carried out
Computing, obtains the first operation result.
In other embodiments, the converting unit 111 can also use the programming different from first operational model
Language realizes the conversion of the data format of the pending data, in this way, the arithmetic element 112 will accord with transformed data
The first input data for closing the input data form of first operational model is inputted into first operational model and transported
During calculation, the input of first input data can be realized by way of calling corresponding data-interface.
Call unit 113 calls data-interface, the defeated of second operational model will be met in the transformed data
The second input data for entering data format inputs and carries out computing into second operational model, obtains the second operation result.
In at least one embodiment of the present invention, will meet in the call unit 113 in the transformed data
During second input data of the input data form of second operational model is inputted into second operational model,
Due to being related to transmission of the data from software runtime environment to hardware running environment, the call unit 113 needs to call
Corresponding data-interface.Specifically, the call unit 113 is by calling dynamic link library to form the data-interface.It is described
Dynamic link library is used for realization the transmission and reception of data, and the call unit 113 calls the dynamic link library can be by institute
The second input data for stating the input data form for meeting second operational model in transformed data is inputted to described
In two operational models.
It should be noted that the present invention realizes that language is not restricted to the dynamic link library, such as：The electronics is set
It is standby the language such as C language, C++ to be used to write the dynamic link library.
In other embodiments, the data-interface can be used for data between the model that different software programming language is realized
Transmission, the present invention be not restricted.
In this way, the call unit 113 makes mutual independence between each operational model, to reach by calling data-interface
Convenient transplanting, convenient extension, the convenient technique effect split, while different running environment is adapted to, and realize to difference
The data calibration of operational model.
In at least one embodiment of the present invention, second operational model, which has, accelerates module, the acceleration module
For accelerating the speed of computing.
Specifically, it is described to accelerate the hardware that module be the acceleration data operation realized by hardware program language
Module, it is in the present embodiment, described to accelerate module to be used for the arithmetic speed for accelerating second operational model.
Specifically, it is described acceleration module can be with FPGA realize or a Software Development Kit, or
It is data operation model under a deep learning frame etc., this is not restricted by the present invention.
Acquiring unit 114 obtains first operation result and second operation result.
In at least one embodiment of the present invention, running environment is mentioned with reference to programming language, the acquiring unit 114 can be with
The acquisition of first operation result and second operation result, this hair are realized by way of calling corresponding data-interface
It is bright not to be restricted.
Second operation result and first operation result is normalized in processing unit 115.
In at least one embodiment of the present invention, in order to eliminate shadow of the different data formats to data results
Ring, second operation result and first operation result is normalized in the processing unit 115, so that described
Second operation result and first operation result are comparable.
Determination unit 116 is according to first operation result after normalized and described second after normalized
Operation result, determines the calibration result of second operational model.
In at least one embodiment of the present invention, the determination unit 116 is according to described first after normalized
Second operation result after operation result and normalized, determining the calibration result of second operational model includes：
The determination unit 116 is by described the after second operation result after normalized and normalized
One operation result is contrasted, and obtains calculation error.
Preferably, when the calculation error is in default error range, the determination unit 116 determines second fortune
It is correct to calculate model；When the calculation error is not in the default error range, the determination unit 116 determines described second
Operational model is inaccurate.
In at least one embodiment of the present invention, the default error range can be according to known in correlative technology field
Empirical value be determined.
In at least one embodiment of the present invention, the determination unit 116 is according to described first after normalized
Second operation result after operation result and normalized, determines that the calibration result of second operational model is also wrapped
Include：
The determination unit 116 using default error analysis method by second operation result after normalized with
First operation result after normalized is contrasted, and obtains the error curve of second operation result, described true
The convergence of the 116 analysis error curve of order member, and when the error curve converges on preset value, the definite list
Member 116 determines that the second operational model computing is correct；It is described to determine when the error curve does not converge on the preset value
Unit 116 determines that the second operational model computing is inaccurate.
In at least one embodiment of the present invention, the default error analysis method includes, but are not limited to：Mean square error
Analysis method, root-mean-square error analysis method, standard deviation analysis method etc., the present invention to the default error analysis method not
It is restricted.
In at least one embodiment of the present invention, the preset value be a less value (such as：1st, 0.89 etc.), this hair
It is bright not to be restricted.
In conclusion the present invention can select pending data from the data set of reception；The pending data is changed
Into the data of at least one data format, the input data form of at least one data format including the first operational model and
The input data form of second operational model, first operational model are used for realization identical calculation with second operational model
Method；First input data of the input data form for meeting first operational model in transformed data is inputted to described
Computing is carried out in first operational model, obtains the first operation result；Data-interface is called, will be met in the transformed data
Second input data of the input data form of second operational model is inputted carries out computing into second operational model,
Obtain the second operation result；Obtain first operation result and second operation result；To second operation result and
First operation result is normalized；According to first operation result and normalized after normalized
Second operation result afterwards, determines the calibration result of second operational model.Therefore, versatility of the present invention is good, and side
Just transplant, convenient extension, conveniently fractionation, be adapted to different running environment, and realize the data school to nonidentity operation model
It is accurate.
As shown in figure 3, it is the structure diagram of the electronic equipment for the preferred embodiment that the present invention realizes data calibration method.
The electronic equipment 1 be it is a kind of can according to the instruction for being previously set or storing, it is automatic carry out numerical computations and/or
The equipment of information processing, its hardware include but not limited to microprocessor, application-specific integrated circuit (Application Specific
Integrated Circuit, ASIC), programmable gate array (Field-Programmable Gate Array, FPGA), number
Word processing device (Digital Signal Processor, DSP), embedded device etc..
The electronic equipment 1 can also be but not limited to any type can with user by keyboard, mouse, remote controler, touch
The mode such as template or voice-operated device carries out the electronic product of human-computer interaction, for example, personal computer, tablet computer, smart mobile phone,
Personal digital assistant (Personal Digital Assistant, PDA), game machine, Interactive Internet TV (Internet
Protocol Television, IPTV), intellectual Wearable etc..
The electronic equipment 1 can also be that the calculating such as desktop PC, notebook, palm PC and cloud server are set
It is standby.
Network residing for the electronic equipment 1 include but not limited to internet, wide area network, Metropolitan Area Network (MAN), LAN, it is virtual specially
With network (Virtual Private Network, VPN) etc..
In one embodiment of the invention, the electronic equipment 1 includes, but not limited to memory 12, processor 13,
And the computer program that can be run in the memory 12 and on the processor 13 is stored in, such as data calibration journey
Sequence.
It will be understood by those skilled in the art that the schematic diagram is only the example of electronic equipment 1, not structure paired electrons
The restriction of equipment 1, can include than illustrating more or fewer components, either combine some components or different components, example
Such as described electronic equipment 1 can also include input-output equipment, network access equipment, bus.
Alleged processor 13 can be central processing unit (Central Processing Unit, CPU), can also be
Other general processors, digital signal processor (Digital Signal Processor, DSP), application-specific integrated circuit
(Application Specific Integrated Circuit, ASIC), ready-made programmable gate array (Field-
Programmable Gate Array, FPGA) either other programmable logic device, discrete gate or transistor logic,
Discrete hardware components etc..General processor can be microprocessor or the processor can also be any conventional processor
Deng the processor 13 is arithmetic core and the control centre of the electronic equipment 1, whole using various interfaces and connection
The various pieces of electronic equipment 1, and perform the operating system of the electronic equipment 1 and types of applications program, the program of installation
Code etc..
The processor 13 performs the operating system of the electronic equipment 1 and the types of applications program of installation.The place
Reason device 13 performs the application program to realize the step in above-mentioned each data calibration method embodiment, such as shown in Fig. 1
Step S10, S11, S12, S13, S14, S15, S16.
Alternatively, the processor 13 realizes each module in above-mentioned each device embodiment/mono- when performing the computer program
The function of member, such as：Pending data is selected from the data set of reception；The pending data is converted at least one number
According to the data of form, at least one data format includes the input data form and the second operational model of the first operational model
Input data form, first operational model and second operational model are used for realization identical algorithm；After changing
Data in meet first operational model the first input data of input data form input to the first computing mould
Computing is carried out in type, obtains the first operation result；Data-interface is called, second fortune will be met in the transformed data
Second input data of the input data form of calculation model is inputted carries out computing into second operational model, obtains the second fortune
Calculate result；Obtain first operation result and second operation result；To second operation result and first fortune
Result is calculated to be normalized；According to described after first operation result after normalized and normalized
Two operation results, determine the calibration result of second operational model.
Exemplary, the computer program can be divided into one or more module/units, one or more
A module/unit is stored in the memory 12, and is performed by the processor 13, to complete the present invention.It is one
Or multiple module/units can be the series of computation machine programmed instruction section that can complete specific function, which is used to retouch
State implementation procedure of the computer program in the electronic equipment 1.For example, the computer program can be divided into choosing
Select unit 110, converting unit 111, arithmetic element 112, call unit 113, acquiring unit 114, processing unit 115, definite list
Member 116, data interface unit 117.
The memory 12 can be used for storing the computer program and/or module, the processor 13 by operation or
The computer program and/or module being stored in the memory 12 are performed, and calls the data being stored in memory 12,
Realize the various functions of the electronic equipment 1.The memory 12 can mainly include storing program area and storage data field, its
In, storing program area can storage program area, application program (such as sound-playing function, image needed at least one function
Playing function etc.) etc.；Storage data field can be stored uses created data (such as voice data, phone directory according to mobile phone
Deng) etc..In addition, memory 12 can include high-speed random access memory, nonvolatile memory can also be included, such as firmly
Disk, memory, plug-in type hard disk, intelligent memory card (Smart Media Card, SMC), secure digital (Secure Digital,
SD) block, flash card (Flash Card), at least one disk memory, flush memory device or other volatile solid-states
Part.
The memory 12 can be the external memory storage and/or internal storage of electronic equipment 1.Further, it is described
Memory 12 can be the circuit with store function for not having in integrated circuit physical form, such as RAM (Random-Access
Memory, random access memory), FIFO (First In First Out) etc..Alternatively, the memory 12 can also be
Memory with physical form, such as memory bar, TF card (Trans-flash Card).
If the integrated module/unit of the electronic equipment 1 is realized in the form of SFU software functional unit and as independent
Production marketing in use, can be stored in a computer read/write memory medium.It is real based on such understanding, the present invention
All or part of flow in existing above-described embodiment method, can also instruct relevant hardware come complete by computer program
Into the computer program can be stored in a computer-readable recording medium, which is being executed by processor
When, it can be achieved that the step of above-mentioned each embodiment of the method.
Wherein, the computer program includes computer program code, and the computer program code can be source code
Form, object identification code form, executable file or some intermediate forms etc..The computer-readable medium can include：Can
Carry any entity or device, recording medium, USB flash disk, mobile hard disk, magnetic disc, CD, the computer of the computer program code
Memory, read-only storage (ROM, Read-Only Memory), random access memory (RAM, Random Access
Memory), electric carrier signal, telecommunication signal and software distribution medium etc..It should be noted that the computer-readable medium
Comprising content appropriate increase and decrease can be carried out according to legislation in jurisdiction and the requirement of patent practice, such as in some departments
Method administrative area, according to legislation and patent practice, computer-readable medium does not include electric carrier signal and telecommunication signal.
With reference to Fig. 1, the memory 12 in the electronic equipment 1 stores multiple instruction to realize a kind of data calibration side
Method, the processor 13 can perform it is the multiple instruction so as to fulfill：Pending data is selected from the data set of reception；By institute
The data that pending data is converted at least one data format are stated, at least one data format includes the first operational model
Input data form and the second operational model input data form, first operational model and second operational model
It is used for realization identical algorithm；First that the input data form of first operational model will be met in transformed data is defeated
Enter data and input and carry out computing into first operational model, obtain the first operation result；Data-interface is called, described will be turned
The second input data for meeting the input data form of second operational model in data after changing is inputted to the described second fortune
Calculate in model and carry out computing, obtain the second operation result；Obtain first operation result and second operation result；To institute
State the second operation result and first operation result is normalized；According to first computing after normalized
As a result second operation result and after normalized, determines the calibration result of second operational model by text to be uploaded
Part is divided into multiple sections；Calculate the cryptographic Hash of first section of the file to be uploaded；Treated when receiving to interrupt described in upload
Upper transmitting file to objective browser trigger signal when, interrupt and upload the file to be uploaded；Record the file to be uploaded
State is uploaded；When receiving when resuming trigger signal of the file to be uploaded, according to first of the file to be uploaded
The cryptographic Hash of section, transfers the file to be uploaded；According to the file to be uploaded of record.
Preferred embodiment according to the present invention, the processor 13, which also performs multiple instruction, to be included：
Total data in the data set of the reception is determined as the pending data；Or
Data in the data set of the reception are ranked up according to the size of data, the data of presetting digit capacity will be come
It is determined as the pending data；Or
The data of preset kind in the data set of the reception are determined as the pending data；Or
The data in the data set of the reception with default mark are determined as the pending data.
Preferred embodiment according to the present invention, first operational model are to realize the algorithm using the first programming language,
Second operational model is to realize the algorithm, first programming language and the described second programming using the second programming language
Language is different.
Preferred embodiment according to the present invention, the processor 13, which also performs multiple instruction, to be included：
By calling dynamic link library to form the data-interface.
Preferred embodiment according to the present invention, second operational model, which has, accelerates module, and the acceleration module is used to add
The speed of fast computing.
Preferred embodiment according to the present invention, the processor 13, which also performs multiple instruction, to be included：
First operation result after second operation result after normalized and normalized is carried out
Contrast, obtains calculation error；
When the calculation error is in default error range, determine that second operational model is correct；Or
When the calculation error is not in the default error range, determine that second operational model is inaccurate.
Preferred embodiment according to the present invention, the processor 13, which also performs multiple instruction, to be included：
Using default error analysis method by after second operation result after normalized and normalized
First operation result is contrasted, and obtains the error curve of second operation result；
Analyze the convergence of the error curve；
When the error curve converges on preset value, determine that the second operational model computing is correct；Or
When the error curve does not converge on the preset value, determine that the second operational model computing is inaccurate.
Specifically, the processor 13 refers to the concrete methods of realizing of above-metioned instruction Fig. 1 and corresponds to correlation in embodiment
The description of step, this will not be repeated here.
In several embodiments provided by the present invention, it should be understood that disclosed system, apparatus and method can be with
Realize by another way.For example, device embodiment described above is only schematical, for example, the module
Division, is only a kind of division of logic function, can there is other dividing mode when actually realizing.
The module illustrated as separating component may or may not be physically separate, be shown as module
The component shown may or may not be physical location, you can with positioned at a place, or can also be distributed to multiple
In network unit.Some or all of module therein can be selected to realize the mesh of this embodiment scheme according to the actual needs
's.
In addition, each function module in each embodiment of the present invention can be integrated in a processing unit, can also
That unit is individually physically present, can also two or more units integrate in a unit.Above-mentioned integrated list
Member can both be realized in the form of hardware, can also be realized in the form of hardware adds software function module.
It is obvious to a person skilled in the art that the invention is not restricted to the details of above-mentioned one exemplary embodiment, Er Qie
In the case of without departing substantially from spirit or essential attributes of the invention, the present invention can be realized in other specific forms.
Therefore, in all respects, the present embodiments are to be considered as illustrative and not restrictive, this
The scope of invention is indicated by the appended claims rather than the foregoing description, it is intended that will fall equivalency in claim
All changes in implication and scope are included in the present invention.Any attached associated diagram mark in claim should not be considered as limit
The involved claim of system.
Furthermore, it is to be understood that one word of " comprising " is not excluded for other units or step, odd number is not excluded for plural number.In system claims
The multiple units or device of statement can also be realized by a unit or device by software or hardware.Second grade word is used
To represent title, and it is not offered as any specific order.
Finally it should be noted that the above embodiments are merely illustrative of the technical solutions of the present invention and it is unrestricted, although reference
The present invention is described in detail in preferred embodiment, it will be understood by those of ordinary skill in the art that, can be to the present invention's
Technical solution is modified or equivalent substitution, without departing from the spirit and scope of technical solution of the present invention.
Claims (10)
- A kind of 1. data calibration method, it is characterised in that the described method includes：Pending data is selected from the data set of reception；The pending data is converted into the data of at least one data format, at least one data format includes first The input data form of operational model and the input data form of the second operational model, first operational model and described second Operational model is used for realization identical algorithm；First input data of the input data form for meeting first operational model in transformed data is inputted to institute State and computing is carried out in the first operational model, obtain the first operation result；Data-interface is called, the second of the input data form of second operational model will be met in the transformed data Input data is inputted carries out computing into second operational model, obtains the second operation result；Obtain first operation result and second operation result；Second operation result and first operation result are normalized；According to first operation result after normalized and second operation result after normalized, institute is determined State the calibration result of the second operational model.
- 2. data calibration method as claimed in claim 1, it is characterised in that selected in the data set from reception pending Data include：Total data in the data set of the reception is determined as the pending data；OrData in the data set of the reception are ranked up according to the size of data, the data for coming presetting digit capacity are determined For the pending data；OrThe data of preset kind in the data set of the reception are determined as the pending data；OrThe data in the data set of the reception with default mark are determined as the pending data.
- 3. data calibration method as claimed in claim 1, it is characterised in that first operational model is using the first programming Language realizes the algorithm, and second operational model is to realize the algorithm, first programming using the second programming language Language is different from second programming language.
- 4. data calibration method as claimed in claim 1, it is characterised in that before calling data-interface, the method further includes：By calling dynamic link library to form the data-interface.
- 5. data calibration method as claimed in claim 1, it is characterised in that second operational model, which has, accelerates module, It is described to accelerate module to be used for the speed for accelerating computing.
- 6. data calibration method as claimed in claim 1, it is characterised in that it is described according to after normalized described first Second operation result after operation result and normalized, determining the calibration result of second operational model includes：Second operation result after normalized and first operation result after normalized are contrasted, Obtain calculation error；When the calculation error is in default error range, determine that second operational model is correct；OrWhen the calculation error is not in the default error range, determine that second operational model is inaccurate.
- 7. data calibration method as claimed in claim 1, it is characterised in that it is described according to after normalized described first Second operation result after operation result and normalized, determines that the calibration result of second operational model is also wrapped Include：Using default error analysis method by described in after second operation result after normalized and normalized First operation result is contrasted, and obtains the error curve of second operation result；Analyze the convergence of the error curve；When the error curve converges on preset value, determine that the second operational model computing is correct；OrWhen the error curve does not converge on the preset value, determine that the second operational model computing is inaccurate.
- 8. a kind of data calibration device, it is characterised in that described device includes：Selecting unit, for selecting pending data from the data set of reception；Converting unit, for the pending data to be converted into the data of at least one data format, at least one number Include the input data form of the first operational model and the input data form of the second operational model, first computing according to form Model is used for realization identical algorithm with second operational model；Arithmetic element, for the first input of the input data form of first operational model will to be met in transformed data Data are inputted carries out computing into first operational model, obtains the first operation result；Call unit, for calling data-interface, will meet the input of second operational model in the transformed data Second input data of data format is inputted carries out computing into second operational model, obtains the second operation result；Acquiring unit, for obtaining first operation result and second operation result；Processing unit, for second operation result and first operation result to be normalized；Determination unit, for according to second fortune after first operation result after normalized and normalized Calculate as a result, determining the calibration result of second operational model.
- 9. a kind of electronic equipment, it is characterised in that the electronic equipment includes：Memory, stores at least one instruction；AndProcessor, performs the instruction that is stored in the memory to realize data as claimed in any of claims 1 to 7 in one of claims Calibration method.
- A kind of 10. computer-readable recording medium, it is characterised in that：At least one is stored with the computer-readable recording medium A instruction, at least one instruction are performed by the processor in electronic equipment to realize such as any one in claim 1 to 7 The data calibration method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711288199.2A CN108038012A (en) | 2017-12-07 | 2017-12-07 | Data calibration method and device, electronic equipment and computer-readable recording medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711288199.2A CN108038012A (en) | 2017-12-07 | 2017-12-07 | Data calibration method and device, electronic equipment and computer-readable recording medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108038012A true CN108038012A (en) | 2018-05-15 |
Family
ID=62095919
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711288199.2A Pending CN108038012A (en) | 2017-12-07 | 2017-12-07 | Data calibration method and device, electronic equipment and computer-readable recording medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108038012A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113454545A (en) * | 2019-01-24 | 2021-09-28 | 西门子股份公司 | Expansion module for independently storing calibration data, assembly and assembly calibration method |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101499937A (en) * | 2009-03-16 | 2009-08-05 | 盛科网络（苏州）有限公司 | Software and hardware collaborative simulation verification system and method based on FPGA |
| CN106528363A (en) * | 2015-09-14 | 2017-03-22 | 深圳市博巨兴实业发展有限公司 | Software and hardware cooperative design verifying method and device |
| CN106970879A (en) * | 2017-03-27 | 2017-07-21 | 南京大学 | A kind of parameter automatization configuration verification platform and method for configurable processing core |
-
2017
- 2017-12-07 CN CN201711288199.2A patent/CN108038012A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101499937A (en) * | 2009-03-16 | 2009-08-05 | 盛科网络（苏州）有限公司 | Software and hardware collaborative simulation verification system and method based on FPGA |
| CN106528363A (en) * | 2015-09-14 | 2017-03-22 | 深圳市博巨兴实业发展有限公司 | Software and hardware cooperative design verifying method and device |
| CN106970879A (en) * | 2017-03-27 | 2017-07-21 | 南京大学 | A kind of parameter automatization configuration verification platform and method for configurable processing core |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113454545A (en) * | 2019-01-24 | 2021-09-28 | 西门子股份公司 | Expansion module for independently storing calibration data, assembly and assembly calibration method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108833458B (en) | Application recommendation method, device, medium and equipment | |
| CN108287857B (en) | Expression picture recommendation method and device | |
| KR20180091043A (en) | Method and apparatus for obtaining user portraits | |
| WO2019024489A1 (en) | Project progress management method, device, terminal apparatus and medium | |
| WO2019148669A1 (en) | Method and apparatus for generating machine learning model, computer device, and storage medium | |
| CN107391526A (en) | A kind of data processing method and equipment based on block chain | |
| CN108363788B (en) | Post intelligent ranking method and device and computer readable storage medium | |
| CN108228861A (en) | For performing the method and system of the Feature Engineering of machine learning | |
| JP6573321B2 (en) | Information processing apparatus, information processing method, and program | |
| CN114708412B (en) | Indoor setting method, device and system based on VR | |
| US10537801B2 (en) | System and method for decision making in strategic environments | |
| WO2021098615A1 (en) | Filling method and device for genotype data missing, and server | |
| CN109542757A (en) | Interface testing environment determines method, apparatus, electronic equipment and storage medium | |
| CN111178537A (en) | Feature extraction model training method and device | |
| CN107016115A (en) | Data export method, device, computer-readable recording medium and electronic equipment | |
| JP2022539290A (en) | Visual programming for deep learning | |
| CN109710747B (en) | Information processing method and device and electronic equipment | |
| CN109656575A (en) | Big data platform quick deployment method, storage medium, electronic equipment and system | |
| CN109934141A (en) | Method and apparatus for labeled data | |
| CN108038012A (en) | Data calibration method and device, electronic equipment and computer-readable recording medium | |
| CN107122401A (en) | To the method for data database storing, equipment, middleware equipment and server | |
| CN109241511B (en) | Electronic report generation method and equipment | |
| CN111191999A (en) | Product research and development management method and device, computer equipment and storage medium | |
| CN109828902A (en) | Interface parameters determines method, apparatus, electronic equipment and storage medium | |
| US10628397B2 (en) | Navigation of data set preparation |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180515 |
|
| RJ01 | Rejection of invention patent application after publication |