# US 2021/0333765 A1 · Environment factor control device and training method thereof

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

- Number: US 2021/0333765 A1
- Document ID: US20210333765A1
- Status: Published application
- Kind: A1
- Application number: 17/238,105
- Filed: 2021-04-22
- Published: 2021-10-28
- Earliest claimed: 2020-04-24
- Assignees: MakinaRocks Co., Ltd.
- Applicants: MakinaRocks Co., Ltd.
- Inventors: Byungchan KIM; Jongsun SHINN; Sangwoo SHIM; Sungho YOON
- Web page: https://patentgrep.com/patent/US20210333765A1

## CPC

- G06N 3/084: Backpropagation, e.g. using gradient descent
- G06N 3/09: Supervised learning
- G05B 13/0265: {the criterion being a learning criterion}
- G05B 13/04: involving the use of models or simulators
- G06F 18/214: Generating training patterns; Bootstrap methods, e.g. bagging or boosting
- G06F 18/24133: {Distances to prototypes}
- G06N 3/008: based on physical entities controlled by simulated intelligence so as to replicate intelligent life forms, e.g. based on robots replicating pets or humans in their appearance or behaviour
- G06N 3/04: Architecture, e.g. interconnection topology
- G06N 3/044: Recurrent networks, e.g. Hopfield networks
- G06N 3/0442: characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
- G06N 3/045: Combinations of networks
- G06N 3/0455: Auto-encoder networks; Encoder-decoder networks
- G06N 3/08: Learning methods
- G06N 3/0985: Hyperparameter optimisation; Meta-learning; Learning-to-learn
- G06N 5/02: Knowledge representation; Symbolic representation
- G06N 20/00: Machine learning
- G06N 3/047: Probabilistic or stochastic 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

Disclosed is a non-transitory computer readable medium storing a computer program, wherein the computer program includes instructions to perform following steps for data processing when the computer program is executed by one or more processors, the steps including: recognizing at least one continuous section from each raw data subset; determining at least one serialization point, based on a start point and an end point of each of the at least one continuous section for each of the raw data subset; and generating a training data set by generating serialized training data, based on the at least one serialization point.

## Claims

### Claim 1

A non-transitory computer readable medium storing a computer program, wherein the computer program includes instructions to perform following steps for data processing when the computer program is executed by one or more processors, the steps comprising:
recognizing at least one continuous section from each raw data subset;
determining at least one serialization point, based on a start point and an end point of each of the at least one continuous section for each of the raw data subset; and
generating training data by generating serialized training data, based on the at least one serialization point.

### Claim 2 (depends on 1)

The non-transitory computer readable medium according to claim 1, wherein the raw data subset is in a form of a step function.

### Claim 3 (depends on 1)

The non-transitory computer readable medium according to claim 1, wherein the serialization point is determined based on a start point and an end point of the continuous section and a predetermined ratio of length of the continuous section.

### Claim 4 (depends on 3)

The non-transitory computer readable medium according to claim 3, wherein the predetermined ratio differs depending on a type of the raw data.

### Claim 5 (depends on 3)

The non-transitory computer readable medium according to claim 3, wherein the serialization point exceeds two within the continuous section.

### Claim 6 (depends on 5)

The non-transitory computer readable medium according to claim 5, wherein the serialization point is determined based on a start point and an end point of two or more subsections for each of the two or more subsections separating the continuous section.

### Claim 7 (depends on 6)

The non-transitory computer readable medium according to claim 6, wherein each of the two or more subsections is different in length.

### Claim 8 (depends on 1)

The non-transitory computer readable medium according to claim 1, wherein the generating of serialized training data is comprised of:
determining a plurality of points connecting the at least one serialization point through an interpolation method.

### Claim 9 (depends on 8)

The non-transitory computer readable medium according to claim 8, wherein the plurality of points is determined by interpolating with a linear functional or a multi-order function.

### Claim 10 (depends on 9)

The non-transitory computer readable medium according to claim 9, wherein the interpolation method is at least one of a linear interpolation or a spline interpolation.

### Claim 11 (depends on 1)

The non-transitory computer readable medium according to claim 1, further comprising:
training an environmental factor control automation model by using the training data set;
evaluating performance of the environmental factor control automation model; and
determining whether to generate a new training data set by resetting the serialization point, based on a result of performance evaluating.

### Claim 12 (depends on 11)

The non-transitory computer readable medium according to claim 11, wherein the performance is measured based on Mean Square Error (MSE) of Validation data set.

### Claim 13

An apparatus for environmental factor control automation, comprising:
a memory; and
a processor;
wherein the processor is configured to:
recognize at least one continuous section from each raw data subset;
determine at least one serialization point, based on a start point and an end point of each of the at least one continuous section for each of the raw data subset; and
generate training data by generating serialized training data, based on the at least one serialization point.

### Claim 14

A non-transitory computer-readable medium storing data structure corresponding to a parameter of neural network where at least a part is updated during a training process, wherein an operation of the neural network is based on at least a part of the parameter, the training process comprising:
recognizing at least one continuous section from each raw data subset;
determining at least one serialization point, based on a start point and an end point of each of the at least one continuous section for each of the raw data subset; and
generating training data by generating serialized training data, based on the at least one serialization point.
