Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Prediction model training method and device and electronic equipment

A prediction model and training method technology, applied in the field of machine learning, can solve the problems of high overhead cost, low prediction accuracy, low multi-value attribute support, etc., and achieve the effect of reducing overhead cost and simplifying deployment steps.

Pending Publication Date: 2020-07-03
BEIJING QIHOO TECH CO LTD
View PDF0 Cites 4 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Therefore, the online processing of the existing DeepFM model has the problem of high overhead cost
In addition, the existing DeepFM model has low support for features of multi-valued attributes, so there is a problem of low prediction accuracy

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Prediction model training method and device and electronic equipment
  • Prediction model training method and device and electronic equipment
  • Prediction model training method and device and electronic equipment

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0068] The embodiment of the present application provides a training method of a prediction model, such as figure 1 As shown, the method includes: step S110 to step S140.

[0069] Step S110, acquiring sample data.

[0070] In this embodiment, the sample data may be obtained through custom editing, or collected from a terminal device. In practical applications, sample data is generally obtained from terminal devices. Specifically, the sample data may include user personal information corresponding to the terminal device, detailed information of applications installed on the terminal device, and the like. In actual application, the user's personal information may include age, gender, hobbies, etc.; the detailed information of the application program may represent the application program installed on the terminal device, such as QQ music, XX security guard, office, etc.

[0071] In practical applications, the terminal device may be electronic devices such as PCs, notebooks, mo...

Embodiment 2

[0114] The embodiment of the present application provides a training device for a predictive model, such as Figure 5 As shown, the device 50 may include: a data acquisition module 501, a vector conversion module 502, a model adjustment module 503 and a model training module 504, wherein,

[0115] A data acquisition module 501, configured to acquire sample data;

[0116] Vector conversion module 502, for utilizing the pre-built DeepFM model to convert sample data into corresponding vector data;

[0117] The model adjustment module 503 is used to obtain the position correspondence of any attribute feature in the vector data in the sample data, and store it in the DeepFM model to obtain the adjusted DeepFM model;

[0118] The model training module 504 is configured to use vector data to train the adjusted DeepFM model to obtain a DeepFM prediction model, so as to deploy the DeepFM prediction model.

[0119] In the embodiment of the present application, the sample data is obtai...

Embodiment 3

[0138] The embodiment of the present application provides an electronic device, such as Figure 6 as shown, Figure 6 The electronic device 600 shown includes: a processor 6001 and a memory 6003 . Wherein, the processor 6001 is connected to the memory 6003 , such as through a bus 6002 . Further, the electronic device 600 may further include a transceiver 6006 . It should be noted that in practical applications, the transceiver 6006 is not limited to one, and the structure of the electronic device 600 does not limit the embodiment of the present application.

[0139] The processor 6001 may be a CPU, a general processor, DSP, ASIC, FPGA or other programmable logic devices, transistor logic devices, hardware components or any combination thereof. It can implement or execute the various illustrative logical blocks, modules and circuits described in connection with the present disclosure. The processor 6001 may also be a combination that implements computing functions, for exam...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The embodiment of the invention provides a prediction model training method and device and electronic equipment. The method comprises the following steps: acquiring sample data; converting the sampledata into corresponding vector data by using a pre-constructed DeepFM model; obtaining a position corresponding relationship of any attribute feature in the sample data in the vector data, and storingthe position corresponding relationship in theDeepFM model to obtain an adjusted DeepFM model; and training the adjusted DeepFM model by using the vector data to obtain a DeepFM prediction model so as to deploy the DeepFM prediction model. According to the embodiment of the invention, the method solves a problem in the prior art that the operation is complex because an index file needs to be built and the format conversion needs to be carried out in an online DeepFM model deployment process, simplifies the deployment steps, and reduces the cost of the online application of the DeepFM prediction model.

Description

technical field [0001] The present application relates to the technical field of machine learning, and in particular, the present application relates to a training method, device and electronic equipment for a predictive model. Background technique [0002] With the development of machine learning algorithms, personalized recommendation functions are widely used. The effect of machine learning algorithms is greatly affected by features, and the work of feature engineering requires a lot of human and material costs, especially the extraction and screening of combined features. In the existing technology, DeepFM is generally used to complete feature extraction and screening. The working principle of DeepFM is: combine deep learning and factorization machine algorithm, and automatically combine first-order, second-order and high-order features, thereby reducing feature engineering. cost and improve the algorithm effect. [0003] The existing DeepFM model has format requiremen...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): G06F16/903G06K9/62G06N3/08
CPCG06N3/08G06F18/214
Inventor 翟羽行
Owner BEIJING QIHOO TECH CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com