Hierarchical latent variable model estimation device, hierarchical latent variable model estimation method, and recording medium

a latent variable and model estimation technology, applied in the field of hierarchical latent variable model estimation device, hierarchical latent variable model estimation method, recording medium, can solve the problems of reducing the marginal likelihood approximation accuracy, algorithm has a problem of extremely high computational complexity, and it is practically impossible to determine the type of observation probability by these two methods

Inactive Publication Date: 2015-03-26
NEC CORP
View PDF1 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Benefits of technology

The present invention provides a device, method, and computer-readable recording medium for estimating a hierarchical latent variable model. This helps solve the problem of selecting the best model for a hierarchical latent variable model and ensures that the model is estimated with appropriate computational complexity without losing its theoretical justification.

Problems solved by technology

This causes a problem of decreased marginal likelihood approximation accuracy.
However, this algorithm has a problem of extremely high computational complexity.
It is practically impossible to determine the type of observation probability by these two methods, due to their problem of computational complexity.

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
  • Hierarchical latent variable model estimation device, hierarchical latent variable model estimation method, and recording medium
  • Hierarchical latent variable model estimation device, hierarchical latent variable model estimation method, and recording medium
  • Hierarchical latent variable model estimation device, hierarchical latent variable model estimation method, and recording medium

Examples

Experimental program
Comparison scheme
Effect test

exemplary embodiment 1

[0053]FIG. 1 is a block diagram showing a structure example of Exemplary Embodiment 1 of a hierarchical latent variable model estimation device according to the present invention. A hierarchical latent variable model estimation device 100 in this exemplary embodiment includes a data input device 101, a hierarchical latent structure setting unit 102, an initialization unit 103, a hierarchical latent variable variational probability computation unit 104, a component optimization unit 105, a gating function optimization unit 106, an optimality determination unit 107, an optimal model selection unit 108, and a model estimation result output device 109.

[0054]The hierarchical latent variable model estimation device 100, upon input of input data 111, optimizes the hierarchical latent structure and the type of observation probability for the input data 111, and outputs the result of optimization as a model estimation result 112.

[0055]FIG. 2 is a block diagram showing a structure example of ...

exemplary embodiment 2

[0105]FIG. 7 is a block diagram showing a structure example of Exemplary Embodiment 2 of a hierarchical latent variable model estimation device according to the present invention. The same structures as those in Exemplary Embodiment 1 are given the same reference signs as in FIG. 1, and their description is omitted. A hierarchical latent variable model estimation device 200 in this exemplary embodiment differs from the hierarchical latent variable model estimation device 100 only in that a hierarchical latent structure optimization unit 201 is connected and the optimal model selection unit 108 is not connected.

[0106]In Exemplary Embodiment 1, the hierarchical latent variable model estimation device 100 optimizes the model of gating functions and components for the hierarchical latent structure candidate, and selects the hierarchical latent structure for optimizing the optimization criterion A. In the hierarchical latent variable model estimation device 200 in this exemplary embodime...

exemplary embodiment 3

[0127]The following describes Exemplary Embodiment 3 of a hierarchical latent variable model estimation device according to the present invention. The hierarchical latent variable model estimation device in this exemplary embodiment differs from the hierarchical latent variable model estimation device 200 only in that the gating function optimization unit 106 is replaced by a gating function optimization unit 113.

[0128]FIG. 11 is a block diagram showing a structure example of the gating function optimization unit 113 in Exemplary Embodiment 3. The gating function optimization unit 113 includes an effective branch node selection unit 113-1 and a branch parameter optimization parallel processing unit 113-2.

[0129]The effective branch node selection unit 113-1 selects only effective branch nodes from the hierarchical latent structure. In detail, through the use of the model 104-5 estimated by the component optimization unit 105, the effective branch node selection unit 113-1 selects onl...

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

A hierarchical latent structure setting unit 81 sets a hierarchical latent structure that is a structure in which latent variables are represented by a tree structure and components representing probability models are located at nodes of a lowest level of the tree structure. A variational probability computation unit 82 computes a variational probability of a path latent variable that is a latent variable included in a path linking a root node to a target node in the hierarchical latent structure. A component optimization unit 83 optimizes each of the components for the computed variational probability. A gating function optimization unit 84 optimizes a gating function model that is a model for determining a branch direction according to the multivariate data in a node of the hierarchical latent structure, on the basis of the variational probability of the latent variable in the node.

Description

CROSS REFERENCE TO RELATED APPLICATIONS[0001]This application is a Continuation application of U.S. application Ser. No. 13 / 758,267 filed Feb. 4, 2013, the disclosures of which are incorporated herein by reference in their entirety.BACKGROUND OF THE INVENTION[0002]1. Field of the Invention[0003]The present invention relates to a hierarchical latent variable model estimation device and a hierarchical latent variable model estimation method for estimating a hierarchical latent variable model for multivariate data, and a computer-readable recording medium having recorded thereon a program for estimating a hierarchical latent variable model for multivariate data.[0004]2. Description of the Related Art[0005]Data typified by sensor data acquired from a car, sales performance of a store, electricity demand history, and the like is data observed according to various factors and accumulated. For example, sensor data acquired from a car varies depending on driving mode. Thus, the data is accu...

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(United States)
IPC IPC(8): G06N7/00
CPCG06N7/005G06F17/18G06N7/01G06F2218/12G06N5/025G06N5/02
Inventor ETO, RIKIFUJIMAKI, RYOHEIMORINAGA, SATOSHI
Owner NEC CORP
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap