CN108805332B - Feature evaluation method and device - Google Patents
Feature evaluation method and device Download PDFInfo
- Publication number
- CN108805332B CN108805332B CN201810427768.5A CN201810427768A CN108805332B CN 108805332 B CN108805332 B CN 108805332B CN 201810427768 A CN201810427768 A CN 201810427768A CN 108805332 B CN108805332 B CN 108805332B
- Authority
- CN
- China
- Prior art keywords
- feature
- evaluation
- evaluated
- target
- preset
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0242—Determining effectiveness of advertisements
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Physics & Mathematics (AREA)
- Game Theory and Decision Science (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Finance (AREA)
- Educational Administration (AREA)
- Accounting & Taxation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The embodiment of the invention provides a feature evaluation method and device, which can receive an evaluation instruction input by a user, wherein the evaluation instruction carries a feature to be evaluated and a data identifier, preprocesses evaluation data corresponding to a target feature according to the data identifier, inputs a preprocessing result of the evaluation data corresponding to the target feature into a pre-trained evaluation model, and determines an evaluation result of the feature to be evaluated according to an output result of the evaluation model. Based on the processing, automatic feature evaluation can be realized, and the efficiency of feature evaluation is further improved.
Description
Technical Field
The present invention relates to the field of data processing technologies, and in particular, to a feature evaluation method and apparatus.
Background
In machine learning, a prediction result can be obtained usually based on a prediction model and raw data. In order to obtain a good prediction result, in addition to selecting a suitable prediction model, the skilled person needs to extract features from the raw data. The quality of the features directly determines the accuracy of the prediction. For example, in predicting the advertisement click rate, the raw data may be data in click requests received by a plurality of advertisements within a preset time period, the data in the click requests may include gender of the user, age of the user, or type of the advertisement, and the technician may use the gender of the user, age of the user, or type of the advertisement as the characteristic for predicting the advertisement click rate.
However, the inventor finds that the prior art has at least the following problems in the process of implementing the invention:
in the prior art, when evaluating the characteristics, a technician often determines the characteristics to be used according to experience. For example, in making a prediction of the advertisement click-through rate, the technician may choose to use the gender of the user as a characteristic and schedule the placement of the advertisement based on the characteristic, and then the technician may determine an evaluation result of the characteristic based on the actual click-through rate of the placed advertisement. The technician may confirm that the feature is valid if there is an increase in the actual click rate of the advertisement, or may confirm that the feature is invalid otherwise.
As can be seen from the above, in the prior art, a technician needs to arrange advertisement delivery according to a certain feature, and after a preset time period, can determine whether a certain feature is valid according to the actual click rate of the advertisement, thereby reducing the efficiency of feature evaluation.
Disclosure of Invention
The embodiment of the invention aims to provide a feature evaluation method and a feature evaluation device so as to improve the efficiency of feature evaluation. The specific technical scheme is as follows:
in a first aspect, to achieve the above object, an embodiment of the present invention discloses a feature evaluation method, where the method includes:
receiving an evaluation instruction input by a user, wherein the evaluation instruction carries features to be evaluated and data identification, the data identification is used for identifying evaluation data required by feature evaluation, and the evaluation data comprises a preset numerical value corresponding to each feature;
according to the data identification, preprocessing evaluation data corresponding to target features, wherein the target features comprise the features to be evaluated and preset features;
inputting a preprocessing result of the evaluation data corresponding to the target feature into a pre-trained evaluation model;
and determining the evaluation result of the feature to be evaluated according to the output result of the evaluation model.
Optionally, the preprocessing the evaluation data corresponding to the target feature according to the data identifier includes:
and if the target feature is a continuous feature, discrete feature transformation is carried out on the evaluation data corresponding to the target feature to obtain a discrete numerical value corresponding to the target feature.
Optionally, the output result of the evaluation model includes a receiver operating characteristic ROC curve, and the determining the evaluation result of the target characteristic according to the output result of the evaluation model includes:
if the area AUC under the ROC curve is larger than a preset threshold value, determining that the characteristic to be evaluated is an effective characteristic;
and if the AUC of the ROC curve is smaller than a preset threshold value, determining that the feature to be evaluated is an invalid feature.
Optionally, when a plurality of features to be evaluated are evaluated, after determining that the features to be evaluated are valid features, the method further includes:
and adding the current feature to be evaluated into the preset feature so as to evaluate the feature of the next feature to be evaluated according to the updated preset feature and the evaluation model.
In a second aspect, to achieve the above object, an embodiment of the present invention discloses a feature evaluation apparatus, including:
the system comprises a receiving and sending module, a judging module and a processing module, wherein the receiving and sending module is used for receiving an evaluation instruction input by a user, the evaluation instruction carries a feature to be evaluated and a data identifier, the data identifier is used for identifying evaluation data required by feature evaluation, and the evaluation data comprises a preset numerical value corresponding to each feature;
the processing module is used for preprocessing evaluation data corresponding to target characteristics according to the data identification, wherein the target characteristics comprise the characteristics to be evaluated and preset characteristics;
inputting a preprocessing result of the evaluation data corresponding to the target characteristics into a pre-trained evaluation model;
and determining the evaluation result of the feature to be evaluated according to the output result of the evaluation model.
Optionally, the processing module is specifically configured to, if the target feature is a continuous feature, perform discrete feature transformation on the evaluation data corresponding to the target feature to obtain a discrete numerical value corresponding to the target feature.
Optionally, the output result of the evaluation model includes a receiver operating characteristic ROC curve, and the processing module is specifically configured to determine that the feature to be evaluated is an effective feature if AUC of the ROC curve is greater than a preset threshold;
and if the AUC of the ROC curve is smaller than a preset threshold value, determining that the feature to be evaluated is an invalid feature.
Optionally, when a plurality of features to be evaluated are evaluated, the processing module is further configured to add a current feature to be evaluated to the preset feature, so as to perform feature evaluation on a next feature to be evaluated according to the updated preset feature and the evaluation model.
In a third aspect, to achieve the above object, an embodiment of the present invention further discloses an electronic device, where the electronic device includes a processor, a communication interface, a memory, and a communication bus, where the processor, the communication interface, and the memory complete communication with each other through the communication bus:
a memory for storing a computer program;
a processor for implementing the method of feature evaluation as described above when executing the program stored in the memory.
In yet another aspect of the present invention, there is also provided a computer-readable storage medium having stored therein instructions, which when executed on a computer, cause the computer to execute any of the above-described feature evaluation methods.
In yet another aspect of the present invention, the embodiment of the present invention further provides a computer program product containing instructions, which when run on a computer, causes the computer to execute any of the above-mentioned feature evaluation methods.
The feature evaluation method and device provided by the embodiment of the invention can receive an evaluation instruction input by a user, wherein the evaluation instruction carries a target feature and a data identifier, preprocesses evaluation data corresponding to the target feature according to the data identifier, inputs a preprocessing result of the evaluation data corresponding to the target feature into a pre-trained evaluation model, and determines an evaluation result of a feature to be evaluated according to an output result of the evaluation model. Based on the processing, automatic feature evaluation can be realized, and the efficiency of feature evaluation is further improved.
Of course, it is not necessary for any product or method of practicing the invention to achieve all of the above-described advantages at the same time.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below.
FIG. 1 is a flow chart of a feature evaluation method according to an embodiment of the present invention;
fig. 2 is a block diagram of a feature evaluation apparatus according to an embodiment of the present invention;
fig. 3 is a structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be described below with reference to the drawings in the embodiments of the present invention.
In the prior art, the characteristics required to be used are often determined empirically by the skilled person. For example, in predicting the click through rate of the advertisement, the technician may choose the gender of the user as a characteristic and arrange the placement of the advertisement according to the characteristic, and then the technician may determine the evaluation result of the characteristic according to the actual click through rate of the placed advertisement. It can be seen that in the prior art, the efficiency of feature evaluation is low.
In order to solve the above problem, embodiments of the present invention provide a feature evaluation method and apparatus, which can be applied to an electronic device, where the electronic device may be a terminal or a server. The electronic equipment can receive an evaluation instruction input by a user, wherein the evaluation instruction carries the features to be evaluated and the data identification. The electronic equipment can preprocess the evaluation data corresponding to the target feature according to the data identification, input a preprocessing result of the evaluation data corresponding to the target feature into a pre-trained evaluation model, and determine an evaluation result of the feature to be evaluated according to an output result of the evaluation model. Based on the processing, the evaluation result of the feature to be evaluated can be directly determined according to the output result of the evaluation model, and the efficiency of feature evaluation can be further improved.
Referring to fig. 1, fig. 1 is a flowchart of a feature evaluation method according to an embodiment of the present invention, where the method may include the following steps:
s101: and receiving an evaluation instruction input by a user.
The evaluation instruction may carry a feature to be evaluated and a data identifier, and the number of the features to be evaluated may be one or multiple. The data identification is used for identifying evaluation data required by feature evaluation, and the evaluation data can comprise a preset numerical value corresponding to each feature. For example, in the prediction of advertisement click-through rates, the evaluation data may be data in click-through requests received for advertisements that have come online.
In practice, the electronic device may typically evaluate a certain feature based on data in the actual business (i.e., evaluation data). For example, when the electronic device evaluates whether a certain feature is effective in predicting the advertisement click rate, the electronic device may obtain click requests received by a plurality of sample advertisements within a certain time period, and use a feature value corresponding to a feature to be evaluated and an online feature included in the click requests as evaluation data. The online feature may be an active feature used in the current service. Then, the electronic device may evaluate the feature to be evaluated according to the feature value corresponding to the feature to be evaluated and the online feature. When a user needs to evaluate a certain feature (i.e., a feature to be evaluated), an evaluation instruction may be input to the electronic device through an input component of the electronic device, and the electronic device may receive the evaluation instruction and analyze the evaluation instruction to obtain the feature to be evaluated and the data identifier.
S102: and preprocessing the evaluation data corresponding to the target characteristics according to the data identification.
The target features may include features to be evaluated and preset features, the preset features may be effective features used in the current service, and the preset features may be one or multiple. Specifically, the preset feature may be preset by a technician according to experience, or may be an effective feature determined by the feature evaluation method according to the embodiment of the present invention.
In implementation, when the electronic device parses the evaluation instruction to obtain the data identifier, evaluation data corresponding to the target feature may be obtained according to the data identifier. Specifically, when the electronic device evaluates whether a certain feature is effective in predicting the advertisement click rate, the electronic device may obtain data (i.e., evaluation data) corresponding to the target feature from data included in a click request of the locally stored advertisement according to the data identifier. For example, data in a click request received for an advertisement may include: male, 24 years old, 9 o' clock in the evening, the user clicks the advertisement by using the mobile phone end, and the clicked advertisement image contains animals. The target features may include: age of the user, gender of the user. The electronic device can obtain the age of the user and the gender of the user in 300 click requests of the locally stored 10 advertisements from 2017, 8/9/2017, 8/10/2017 according to the data identification. Or, the electronic device may also obtain data included in click requests of a preset number of advertisements received by a server providing an advertisement service within a certain time period, and extract evaluation data corresponding to the target features. The electronic device may then pre-process the evaluation data corresponding to the target feature. Specifically, the method for preprocessing the evaluation data by the electronic device will be described in detail later.
S103: and inputting the preprocessing result of the evaluation data corresponding to the target characteristics into a pre-trained evaluation model.
The evaluation model may be an FM (factor Machine) model, an LR (Logistic Regression) model, or other evaluation models in the prior art. The evaluation model may be obtained by training evaluation data corresponding to the valid features that have been online, for example, the evaluation model may be obtained by training an FM model according to the preset features and the actual click rate of the sample advertisement.
In implementation, after the electronic device preprocesses the evaluation data corresponding to the target feature, the preprocessing result may be input to a pre-trained evaluation model to evaluate the feature to be evaluated.
S104: and determining the evaluation result of the feature to be evaluated according to the output result of the evaluation model.
The output result of the evaluation model may be an ROC (Receiver Operating characteristics) curve, and may also be a PR (Precision Recall accuracy) curve.
In implementation, the electronic device may determine whether the feature to be evaluated is valid according to an output result of the evaluation model. Specifically, it will be described in detail later.
As can be seen from the above, according to the feature evaluation method provided by the embodiment of the present invention, the electronic device may receive an evaluation instruction input by a user, preprocess evaluation data corresponding to a target feature according to a data identifier, input a preprocessing result to a pre-trained evaluation model, and determine an evaluation result of a feature to be evaluated according to an output result of the evaluation model. Based on the processing, the evaluation result of the feature to be evaluated can be directly determined according to the output result of the evaluation model, and therefore the efficiency of feature evaluation can be improved.
Optionally, the electronic device may perform preprocessing on the evaluation data corresponding to the target feature according to the data identifier, and the preprocessing may include the following steps: and if the target feature is a continuous feature, performing discrete feature transformation on the evaluation data corresponding to the target feature to obtain a discrete numerical value corresponding to the target feature.
In an implementation, if the target feature is a continuous feature, the electronic device may perform discrete feature transformation on the evaluation data corresponding to the target feature to obtain a discrete numerical value corresponding to the target feature, for example, the target feature may be an age, the evaluation data corresponding to the target feature includes 20 years to 29 years, and 10 ages are counted, and the electronic device may take each age as one feature and perform numbering, where the numbering is 1 to 10. If the data included in a certain click request is 20 years old, the electronic device may determine that the value on the feature of the click request number 1 is 1 and the values on the features of numbers 2 to 10 are 0, and correspondingly, if the data included in a certain click request is 25 years old, the electronic device may determine that the value on the feature of the click request number 6 is 1 and the values on the features of numbers 1 to 5 and 7 to 10 are 0. In addition, when the target feature further includes gender, the electronic device may take each gender as one feature and perform numbering, with numbers 11 and 12, the number 11 representing male, and the number 12 representing female. If the data included in a certain click request is male, the electronic device may determine that the value of the feature of the click request number 11 is 1 and the value of the feature of the click request number 12 is 0, and correspondingly, if the data included in a certain click request is female, the electronic device may determine that the value of the feature of the click request number 10 is 0 and the value of the feature of the click request number 12 is 1. Specifically, the electronic device may perform discrete feature transformation on the continuous feature according to a GBDT (Gradient Boosting Decision Tree). Furthermore, the electronic device may obtain a discrete numerical value corresponding to the target feature in the click request.
In addition, in the prediction of the click rate of the advertisement, it is generally necessary to acquire the image characteristics of the advertisement, and the image characteristics may include the number of faces included in a picture of the advertisement, the number of cars included in a picture of the advertisement, and the like. The electronic equipment can acquire image data corresponding to the target feature according to a preset image feature extraction algorithm, and then perform discrete feature transformation on the image data to obtain a discrete numerical value corresponding to the target feature. For example, the target feature is the number of faces contained in the image, and for a certain click request, the electronic device may obtain an image corresponding to the click request, then obtain the number of faces contained in the image according to an image feature extraction algorithm, and use the number of faces contained in the image as image data corresponding to the target feature. Then, the electronic device may perform discrete feature transformation on the image data to obtain a discrete numerical value corresponding to the target feature. Specifically, reference may be made to detailed description of discrete feature transformation performed on evaluation data corresponding to a target feature by the electronic device in the foregoing embodiment.
As can be seen from the above, based on the feature evaluation method of the embodiment of the present invention, the electronic device may automatically pre-process the evaluation data corresponding to the target feature to obtain a discrete value corresponding to the target feature, so that the efficiency of feature evaluation can be improved.
Optionally, the output result of the evaluation model may include an ROC curve, and the electronic device determines the evaluation result of the target feature according to the output result of the evaluation model, and may include the following steps: if the AUC of the ROC curve is larger than a preset threshold value, determining the characteristic to be evaluated as an effective characteristic; and if the AUC of the ROC curve is smaller than a preset threshold value, determining that the characteristic to be evaluated is an invalid characteristic.
Wherein the preset threshold value can be set by a technician according to experience.
In implementation, the electronic device may obtain an output result of the evaluation model, specifically, the electronic device may evaluate an ROC curve output by the model, and calculate an AUC of the ROC curve, and then the electronic device may determine whether the AUC of the ROC curve is greater than a preset threshold, when the electronic device determines that the AUC of the ROC curve is greater than or equal to the preset threshold, the electronic device may determine that the feature to be evaluated is an effective feature, and when the electronic device determines that the AUC of the ROC curve is less than the preset threshold, the electronic device may determine that the feature to be evaluated is an ineffective feature.
As can be seen from the above, based on the feature evaluation method of the embodiment of the present invention, the electronic device can directly determine whether the feature to be evaluated is valid according to the AUC of the ROC curve output by the evaluation model, and the efficiency of feature evaluation can be improved.
Optionally, when the electronic device evaluates a plurality of features to be evaluated, after determining that the current feature to be evaluated is a valid feature, the method may further include the following steps: and adding the current feature to be evaluated into the preset feature so as to evaluate the feature of the next feature to be evaluated according to the updated preset feature and the evaluation model.
In implementation, after the electronic device determines that the current feature to be evaluated is an effective feature, the electronic device may use the feature to be evaluated as a preset feature, so that the electronic device may perform feature evaluation on the next feature to be evaluated according to the updated preset feature and a pre-trained evaluation model, thereby improving accuracy and efficiency of feature evaluation.
Corresponding to the embodiment of the method in fig. 1, referring to fig. 2, fig. 2 is a block diagram of a feature evaluation apparatus according to an embodiment of the present invention, where the apparatus may include:
the system comprises a transceiver module 201, a processing module and a processing module, wherein the transceiver module is used for receiving an evaluation instruction input by a user, the evaluation instruction carries a feature to be evaluated and a data identifier, the data identifier is used for identifying evaluation data required by feature evaluation, and the evaluation data comprises a preset numerical value corresponding to each feature;
the processing module 202 is configured to perform preprocessing on evaluation data corresponding to a target feature according to the data identifier, where the target feature includes the feature to be evaluated and a preset feature;
inputting a preprocessing result of the evaluation data corresponding to the target characteristics into a pre-trained evaluation model;
and determining the evaluation result of the feature to be evaluated according to the output result of the evaluation model.
Optionally, the processing module 202 is specifically configured to, if the target feature is a continuous feature, perform discrete feature transformation on the evaluation data corresponding to the target feature to obtain a discrete numerical value corresponding to the target feature.
Optionally, the output result of the evaluation model includes a receiver operating characteristic ROC curve, and the processing module is specifically configured to determine that the feature to be evaluated is an effective feature if AUC of the ROC curve is greater than a preset threshold;
and if the AUC of the ROC curve is smaller than a preset threshold value, determining that the feature to be evaluated is an invalid feature.
Optionally, when a plurality of features to be evaluated are evaluated, the processing module 202 is further configured to add a current feature to be evaluated to the preset feature, so as to perform feature evaluation on a next feature to be evaluated according to the updated preset feature and the evaluation model.
As can be seen from the above, based on the feature evaluation apparatus in the embodiment of the present invention, the electronic device may receive an evaluation instruction input by a user, preprocess evaluation data corresponding to a target feature according to a data identifier, input a preprocessing result to a pre-trained evaluation model, and determine an evaluation result of a feature to be evaluated according to an output result of the evaluation model. Based on the above processing, the efficiency of feature evaluation can be improved.
An embodiment of the present invention further provides an electronic device, as shown in fig. 3, including a processor 301, a communication interface 302, a memory 303 and a communication bus 304, where the processor 301, the communication interface 302 and the memory 303 complete mutual communication through the communication bus 304,
a memory 303 for storing a computer program;
the processor 301 is configured to implement the following steps when executing the program stored in the memory 303:
receiving an evaluation instruction input by a user, wherein the evaluation instruction carries a feature to be evaluated and a data identifier, the data identifier is used for identifying evaluation data required by feature evaluation, and the evaluation data comprises a preset numerical value corresponding to each feature;
according to the data identification, preprocessing evaluation data corresponding to target features, wherein the target features comprise the features to be evaluated and preset features;
inputting a preprocessing result of the evaluation data corresponding to the target feature into a pre-trained evaluation model;
and determining the evaluation result of the feature to be evaluated according to the output result of the evaluation model.
The communication bus 304 mentioned in the above electronic device may be a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The communication bus 304 may be divided into an address bus, a data bus, a control bus, and the like. For ease of illustration, only one thick line is shown, but this is not intended to represent only one bus or type of bus.
The communication interface 302 is used for communication between the above-described electronic apparatus and other apparatuses.
The Memory 303 may include a Random Access Memory (RAM), and may also include a non-volatile Memory (non-volatile Memory), such as at least one disk Memory. Optionally, the memory 303 may also be at least one storage device located remotely from the processor 301.
The Processor 301 may be a general-purpose Processor, and includes a Central Processing Unit (CPU), a Network Processor (NP), and the like; the Integrated Circuit may also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other Programmable logic device, a discrete Gate or transistor logic device, or a discrete hardware component.
In yet another embodiment of the present invention, a computer-readable storage medium is further provided, which has instructions stored therein, and when the computer-readable storage medium runs on a computer, the computer is caused to execute the feature evaluation method described in any one of the above embodiments.
In yet another embodiment, there is provided a computer program product comprising instructions which, when run on a computer, cause the computer to perform the method of feature assessment described in any of the above embodiments.
In the above embodiments, the implementation may be wholly or partially realized by software, hardware, firmware, or any combination thereof. When implemented in software, it may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When loaded and executed on a computer, cause the processes or functions described in accordance with the embodiments of the invention to be performed in whole or in part. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable device. The computer instructions may be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another computer readable storage medium, for example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital Subscriber Line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device, such as a server, a data center, etc., that includes one or more available media. The usable medium may be a magnetic medium (e.g., floppy Disk, hard Disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid State Disk (SSD)), among others.
It should be noted that, in this document, relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrases "comprising a," "8230," "8230," or "comprising" does not exclude the presence of additional like elements in a process, method, article, or apparatus that comprises the element.
All the embodiments in the present specification are described in a related manner, and the same and similar parts among the embodiments may be referred to each other, and each embodiment focuses on differences from other embodiments. In particular, for the apparatus, the electronic device, the computer-readable storage medium, and the computer program product embodiments, since they are substantially similar to the method embodiments, the description is relatively simple, and for the relevant points, reference may be made to the partial description of the method embodiments.
Claims (9)
1. A method of feature evaluation, the method comprising:
receiving an evaluation instruction input by a user, wherein the evaluation instruction carries a feature to be evaluated and a data identifier, the data identifier is used for identifying evaluation data required by feature evaluation, and the evaluation data comprises a preset numerical value corresponding to each feature;
according to the data identification, preprocessing evaluation data corresponding to target features, wherein the target features comprise the features to be evaluated and preset features;
inputting a preprocessing result of the evaluation data corresponding to the target characteristics into a pre-trained evaluation model;
and determining an evaluation result of the feature to be evaluated according to an output result of the evaluation model, wherein the evaluation result is used for judging whether the feature to be evaluated is a valid feature.
2. The method according to claim 1, wherein the preprocessing the evaluation data corresponding to the target feature according to the data identifier comprises:
and if the target feature is a continuous feature, performing discrete feature transformation on the evaluation data corresponding to the target feature to obtain a discrete numerical value corresponding to the target feature.
3. The method of claim 1, wherein the output of the evaluation model comprises a Receiver Operating Characteristic (ROC) curve, and wherein determining the evaluation of the target feature based on the output of the evaluation model comprises:
if the area AUC under the ROC curve is larger than a preset threshold value, determining that the feature to be evaluated is an effective feature;
or if the AUC of the ROC curve is smaller than a preset threshold value, determining that the feature to be evaluated is an invalid feature.
4. The method of claim 3, wherein when evaluating a plurality of features to be evaluated, after determining that the features to be evaluated are valid features, the method further comprises:
and adding the current feature to be evaluated into the preset feature so as to evaluate the feature of the next feature to be evaluated according to the updated preset feature and the evaluation model.
5. A feature evaluation apparatus, characterized in that the apparatus comprises:
the system comprises a receiving and sending module, a judging module and a processing module, wherein the receiving and sending module is used for receiving an evaluation instruction input by a user, the evaluation instruction carries a feature to be evaluated and a data identifier, the data identifier is used for identifying evaluation data required by feature evaluation, and the evaluation data comprises a preset numerical value corresponding to each feature;
the processing module is used for preprocessing evaluation data corresponding to target characteristics according to the data identification, wherein the target characteristics comprise the characteristics to be evaluated and preset characteristics;
inputting a preprocessing result of the evaluation data corresponding to the target feature into a pre-trained evaluation model;
and determining an evaluation result of the feature to be evaluated according to an output result of the evaluation model, wherein the evaluation result is used for judging whether the feature to be evaluated is an effective feature.
6. The apparatus of claim 5,
the processing module is specifically configured to perform discrete feature transformation on the evaluation data corresponding to the target feature to obtain a discrete numerical value corresponding to the target feature if the target feature is a continuous feature.
7. The apparatus according to claim 5, wherein the output of the evaluation model comprises a Receiver Operating Characteristic (ROC) curve, and the processing module is specifically configured to determine that the feature to be evaluated is a valid feature if an area under the curve (AUC) of the ROC curve is greater than a preset threshold;
and if the AUC of the ROC curve is smaller than a preset threshold value, determining that the feature to be evaluated is an invalid feature.
8. The apparatus of claim 7, wherein when a plurality of features to be evaluated are evaluated, the processing module is further configured to add a current feature to be evaluated to the preset feature, so as to evaluate a feature of a next feature to be evaluated according to the updated preset feature and the evaluation model.
9. An electronic device, comprising a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus;
the memory is used for storing a computer program;
the processor, when executing the program stored in the memory, implementing the method steps of any of claims 1-4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810427768.5A CN108805332B (en) | 2018-05-07 | 2018-05-07 | Feature evaluation method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810427768.5A CN108805332B (en) | 2018-05-07 | 2018-05-07 | Feature evaluation method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108805332A CN108805332A (en) | 2018-11-13 |
| CN108805332B true CN108805332B (en) | 2022-12-02 |
Family
ID=64092002
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810427768.5A Active CN108805332B (en) | 2018-05-07 | 2018-05-07 | Feature evaluation method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108805332B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109858633B (en) * | 2019-02-22 | 2021-02-02 | 中国工商银行股份有限公司 | Characteristic information identification method and system |
| CN110188942A (en) * | 2019-05-27 | 2019-08-30 | 北京金山安全软件有限公司 | Click through rate prediction method, device and equipment |
| CN110189171B (en) * | 2019-05-27 | 2021-08-06 | 北京金山安全软件有限公司 | Feature data generation method, device and equipment |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105610929A (en) * | 2015-12-24 | 2016-05-25 | 北京奇虎科技有限公司 | Personalized data pushing method and device |
| CN106951925A (en) * | 2017-03-27 | 2017-07-14 | 成都小多科技有限公司 | Data processing method, device, server and system |
| CN107844867A (en) * | 2017-11-22 | 2018-03-27 | 广州优视网络科技有限公司 | Data processing method, device and equipment |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8069076B2 (en) * | 2003-03-25 | 2011-11-29 | Cox Communications, Inc. | Generating audience analytics |
| CN107193805B (en) * | 2017-06-06 | 2021-05-14 | 北京百度网讯科技有限公司 | Article value evaluation method and device based on artificial intelligence and storage medium |
-
2018
- 2018-05-07 CN CN201810427768.5A patent/CN108805332B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105610929A (en) * | 2015-12-24 | 2016-05-25 | 北京奇虎科技有限公司 | Personalized data pushing method and device |
| CN106951925A (en) * | 2017-03-27 | 2017-07-14 | 成都小多科技有限公司 | Data processing method, device, server and system |
| CN107844867A (en) * | 2017-11-22 | 2018-03-27 | 广州优视网络科技有限公司 | Data processing method, device and equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108805332A (en) | 2018-11-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107305611B (en) | Method and device for establishing model corresponding to malicious account and method and device for identifying malicious account | |
| CN108805332B (en) | Feature evaluation method and device | |
| CN110728323A (en) | Target type user identification method and device, electronic equipment and storage medium | |
| WO2020257991A1 (en) | User identification method and related product | |
| CN111080374B (en) | Advertisement putting strategy testing method, bidding server and advertisement putting system | |
| WO2023000491A1 (en) | Application recommendation method, apparatus and device, and computer-readable storage medium | |
| CN112669078A (en) | Behavior prediction model training method, device, equipment and storage medium | |
| CN111539780A (en) | Task processing method and device, storage medium and electronic equipment | |
| CN114330550A (en) | Model fusion strategy selection method and device and electronic equipment | |
| CN108647986B (en) | Target user determination method and device and electronic equipment | |
| CN108021713B (en) | Document clustering method and device | |
| CN106817296B (en) | Information recommendation test method and device and electronic equipment | |
| CN108667875B (en) | Information updating method and device | |
| CN111404835B (en) | Flow control method, device, equipment and storage medium | |
| CN110996142B (en) | Video recall method and device, electronic equipment and storage medium | |
| CN113205189B (en) | Method for training prediction model, prediction method and device | |
| CN111080349B (en) | Method, device, server and medium for identifying multiple devices of same user | |
| CN110704614B (en) | Information processing method and device for predicting user group type in application | |
| CN110717109B (en) | Method, device, electronic equipment and storage medium for recommending data | |
| CN110569346B (en) | Automatic service processing method, device, equipment and storage medium | |
| CN110020728B (en) | Service model reinforcement learning method and device | |
| CN113536138A (en) | Network resource recommendation method and device, electronic equipment and readable storage medium | |
| CN112200602A (en) | Neural network model training method and device for advertisement recommendation | |
| CN116383664A (en) | Satisfaction model training and satisfaction determining method and device and electronic equipment | |
| CN109064216B (en) | Method and device for simulating exposure of advertisement order |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |