# US 2021/0341887 A1 · Information processing device and information processing method

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2021/0341887 A1
- Document ID: US20210341887A1
- Status: Published application
- Kind: A1
- Application number: 17/186,233
- Filed: 2021-02-26
- Published: 2021-11-04
- Earliest claimed: 2020-04-30
- Assignees: Hitachi, Ltd.
- Applicants: Hitachi, Ltd.
- Inventors: Yuya OKADOME; Toshiko AIZONO
- Web page: https://patentgrep.com/patent/US20210341887A1

## CPC

- G06N 20/00: Machine learning
- G05B 13/0265: {the criterion being a learning criterion}
- G05B 13/042: {in which a parameter or coefficient is automatically adjusted to optimise the performance}
- G06F 18/2148: {characterised by the process organisation or structure, e.g. boosting cascade}
- G06F 18/2415: based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
- G06N 5/01: Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
- G06N 7/01: Probabilistic graphical models, e.g. probabilistic networks
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G05B: CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS (systems for controlling or regulating non-electric variables G05D; systems for regulating electric or magnetic variables G05F; control devices or systems insofar as characterised by mechanical features only G05G)
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

To provide an information processing device capable of more efficiently searching for parameters. An optimization portion selects a search area, namely, any of the multiple subspaces contained in a parameter space composed of multiple parameters based on evaluation result information indicating an evaluation value that evaluates a processing result of information processing based on each trial parameter by using the trial parameter. The optimization portion allows a classification model portion to perform information processing using the trial parameter, namely, any one of parameters belonging to the search area. The optimization portion repeats a search process to update the evaluation result information based on the processing result of the information processing.

## Claims

### Claim 1

An information processing device to search for an available parameter used for predetermined information processing, comprising:
an information processing portion to perform the information processing; and
a search portion that selects a search area, namely, any of a plurality of subspaces included in a parameter space comprised of a plurality of parameters based on evaluation result information indicating an evaluation value to evaluate a processing result of the information processing based on each trial parameter by using the trial parameter, allows the information processing portion to perform the information processing by using any of parameters belonging to the search area as the trial parameter, and repeats a search process to update the evaluation result information based on a processing result of the information processing.

### Claim 2 (depends on 1)

The information processing device according to claim 1,
wherein the search portion provides a selection probability for each of the subspaces based on the evaluation result information and selects the search area, according to the selection probability.

### Claim 3 (depends on 2)

The information processing device according to claim 2,
wherein, the search portion calculates an aggregate value for each of the subspaces based on the evaluation result information, the aggregate value being configured to aggregate evaluation values corresponding to trial parameters belonging to the subspace, and provides the selection probability based on the aggregate value.

### Claim 4 (depends on 3)

The information processing device according to claim 3,
wherein the search portion provides the selection probability based on the number of searches to select the search area in each subspace.

### Claim 5 (depends on 1)

The information processing device according to claim 1,
wherein the subspace is generated by dividing an order-reduced space resulting from order reduction of the parameter space.

### Claim 6 (depends on 1)

The information processing device according to claim 1,
wherein the search portion repeats the search process a predetermined number of times and then determines the available parameter based on the evaluation result information.

### Claim 7 (depends on 1)

The information processing device according to claim 1,
wherein the search portion outputs a display screen based on the evaluation result information; and
wherein, for each of the subspaces, the display screen shows evaluation information comparable to an evaluation value corresponding to a trial parameter belonging to the subspace.

### Claim 8 (depends on 1)

The information processing device according to claim 1,
wherein the information processing is comparable to machine learning.

### Claim 9 (depends on 8)

The information processing device according to claim 8,
wherein the parameter is comparable to a hyperparameter.

### Claim 10 (depends on 8)

The information processing device according to claim 8,
wherein the parameter is comparable to a feature quantity.

### Claim 11

An information processing method of searching for an available parameter used for predetermined information processing performed by an information processing device, comprising the step of:
selecting a search area, namely, any of a plurality of subspaces included in a parameter space comprised of a plurality of parameters based on evaluation result information indicating an evaluation value to evaluate a processing result of the information processing based on each trial parameter by using the trial parameter, performing the information processing by using any of parameters belonging to the search area as the trial parameter, and repeating a search process to update the evaluation result information based on a processing result of the information processing.
