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

High-order factor decomposer method based on sparse group Lasso

A factoring, high-level technology, applied in instruments, complex mathematical operations, calculations, etc., can solve problems such as the inability to directly use the prior information of the cross-term model, and achieve the effect of good performance, improved noise characteristics, and fewer parameters

Inactive Publication Date: 2019-08-20
NANJING UNIV OF AERONAUTICS & ASTRONAUTICS
View PDF2 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0014] The current typical and popular method of feature selection is by adding The regularization term of the norm is implemented, so that although sparse solutions can be obtained, for high-order factorization machine models, The norm cannot directly use the prior information of the cross-term model

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
  • High-order factor decomposer method based on sparse group Lasso
  • High-order factor decomposer method based on sparse group Lasso
  • High-order factor decomposer method based on sparse group Lasso

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0035] 1. The special high-order cross structure of the high-order factorization machine is used, and the SGL regular term is used to punish the objective function, thereby realizing the structural sparseness of the feature, thereby removing noise and improving the performance of the model. The objective function of the new model is as follows:

[0036]

[0037] 2. Use the FOBOS algorithm to optimize the new objective function. The FOBOS algorithm is an iterative optimization algorithm. The specific process is as follows:

[0038] According to the FOBOS algorithm, formula (7) is first divided into two parts

[0039] f(ω,p (2) ,...,p (m) )+r(ω,p (2) ,...,p (m) ) (8)

[0040] in and The FOBOS update process is:

[0041] Step 1: Choose an appropriate regularization parameter β 1 and beta 2 , and the learning rate η, let time t=0.

[0042] Step 2: Yes Perform the standard gradient descent procedure.

[0043] Step 3: On the basis of Step 2, use the optimization al...

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 invention relates to an enhancement strategy of a high-order factor decomposer model, in particular to a high-order factor decomposer method (simply written as SGL-HOFM) based on sparse group Lasso. The high-order factor decomposer method is characterized in that a Sparse Group Lasso-based regular item is adopted for a special intersection item structure of a high-order factor decomposition machine instead of a traditional two-norm, and a new model can realize sparse structure of features and play a role in feature selection. After the high-order factor decomposer method is adopted, the number of parameters of the model is obviously reduced, so that the required storage space is reduced; due to the fact that noise is removed through feature selection, the performance of the model is improved; and the high-order factor decomposer method can be widely applied to the field of data mining, such as classification, regression, sorting, feature analysis, a recommendation system and socialnetwork analysis, and is particularly suitable for solving the analysis problem containing large-scale sparse high-dimensional data.

Description

[0001] 1. Technical field [0002] The present invention is an enhancement strategy of a high-order factorization machine model, specifically "a high-order factorization machine method based on sparse group Lasso" (abbreviated as SGL-HOFM), which is mainly used in the field of data mining, such as classification, regression, sorting, Various scenarios such as feature analysis, recommendation system, and social network analysis are especially suitable for dealing with analysis problems involving large-scale sparse high-dimensional data. [0003] 2. Background technology [0004] Factorization Machine (Factorization Machine, FM for short) is a machine learning algorithm proposed by Steffen Rendle in 2010, which mainly solves the problem of feature combination under large-scale sparse data, so as to be used in applications such as data mining and recommendation systems. The main task of the latter is to predict the items that the user may purchase in the future based on the user's...

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
Patent Type & Authority Applications(China)
IPC IPC(8): G06F17/15
CPCG06F19/00
Inventor 陈松灿郭少成
Owner NANJING UNIV OF AERONAUTICS & ASTRONAUTICS
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