IP Library Granted Patent US 12675736
Granted Patent B2
US 12675736 · App. 18/103,759 · Granted Jul 7, 2026

Machine learning method for predicting a sensory result

Inventor: Keiko Matsumoto (Kyoto, JP)
Assignee: SHIMADZU CORPORATION
G06N20/00
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12675736
App. No.
18/103,759
Granted
Jul 7, 2026
Kind
B2
Abstract

A second device performs receiving an analysis result of analyzing a food sample by an analysis instrument, examining the analysis result to obtain an examination result, acquiring first information regarding whether the analysis result is made public or private and second information regarding whether the examination result is made public or private, storing the first information and the analysis result in association with each other and storing the second information and the examination result in association with each other, and performing machine learning for predicting an examination result using, as training data, at least either the analysis result or the examination result, the analysis result or the examination result being set public.

Claims (28)

1 . A machine learning method that is performed by a computing device belonging to a system management company, and a computing device of a client,

the machine learning method causing the computing device belonging to the system management company to perform:

acquiring an analysis result of analyzing a sample by an analysis instrument and a first sensory result of the sample;

acquiring public setting information regarding whether the analysis result and the first sensory result are made public or private;

storing the public setting information, the analysis result, and the first sensory result, the public setting information being associated with the analysis result and the first sensory result;

creating a learned model by performing machine learning for predicting the first sensory result using, as training data, at least either the analysis result or the first sensory result, the analysis result or the first sensory result being set public; and

outputting a second sensory result by inputting the analysis result that is set private to the learned model, and

the machine learning method causing the computing device of the client to perform;

setting the public setting information to individually make each of the analysis result and the first sensory result public or private; and

transmitting the public setting information to the computing device belonging to the system management company.

2 . The machine learning method according to claim 1 , wherein

the first sensory result is any one of a sensory evaluation result of the sample, a sensory prediction result output by the computing device belonging to the system management company based on the analysis result, or a study result associated with the analysis result and the sensory evaluation result or the analysis result and the sensory prediction result.

3 . The machine learning method according to claim 2 , wherein

the sensory evaluation result is a result obtained by evaluating the sample by a sensory measurement evaluator.

4 . The machine learning method according to claim 2 , further causing the computing device belonging to the system management company to perform carrying out an operation on the analysis result input and outputting the sensory prediction result.

5 . The machine learning method according to claim 2 , further causing the computing device belonging to the system management company to perform carrying out an operation on the analysis result, and the sensory evaluation result or the sensory prediction result and outputting the study result.

6 . The machine learning method according to claim 1 , wherein

the public setting information is stored in association with client information belonging to the system management company.

7 . The machine learning method according to claim 6 , further causing the computing device belonging to the system management company to perform changing information regarding an amount charged to a client in accordance with the public setting information associated with the client information.

8 . The machine learning method according to claim 1 , further causing the computing device belonging to the system management company to perform making a setting so as to make the analysis result, the first sensory result, and the second sensory result public on a server via a network, the analysis result, the first sensory result, and the second sensory result being set public in the public setting information, and to make the analysis result, the first sensory result, and the second sensory result private on the server, the analysis result, the first sensory result, and the second sensory result being set private in the public setting information.

9 . The machine learning method according to claim 1 , wherein the public setting information includes

first information regarding whether the analysis result is made public or private, and

second information regarding whether the first sensory result and the second sensory result are made public or private.

10 . The machine learning method according to claim 1 ,

further causing the computing device belonging to the system management company to perform transmitting the analysis result and the first sensory result to the computing device of the client, and

further causing the computing device of the client to receive the analysis result and the first sensory result from the computing device belonging to the system management company.

11 . The machine learning method according to claim 1 , wherein the acquired public setting information is received from a customer computer that individually sets the acquired public setting information for the analysis result as public or private.

12 . The machine learning method according to claim 1 , wherein the acquired analysis result includes data provided by multiple clients.