18291238. LEARNING SYSTEM, LEARNING SERVER APPARATUS, PROCESSING APPARATUS, LEARNING METHOD, AND PROGRAM simplified abstract (NIPPON TELEGRAPH AND TELEPHONE CORPORATION)

LEARNING SYSTEM, LEARNING SERVER APPARATUS, PROCESSING APPARATUS, LEARNING METHOD, AND PROGRAM

Organization Name

NIPPON TELEGRAPH AND TELEPHONE CORPORATION

Inventor(s)

Iifan Tyou of Musashino-shi, Tokyo (JP)

Takumi Fukami of Musashino-shi, Tokyo (JP)

LEARNING SYSTEM, LEARNING SERVER APPARATUS, PROCESSING APPARATUS, LEARNING METHOD, AND PROGRAM - A simplified explanation of the abstract

This abstract first appeared for US patent application 18291238 titled 'LEARNING SYSTEM, LEARNING SERVER APPARATUS, PROCESSING APPARATUS, LEARNING METHOD, AND PROGRAM

The learning system described in the abstract significantly reduces the search time of Neural Architecture Search (NAS) and makes machine learning more practical.

• Learning system includes a learning server apparatus and multiple processing apparatuses.
• Processing apparatus calculates a score when local data is applied to neural networks.
• Learning server apparatus aggregates neural networks using scores and selects an optimal neural network.
• Federated learning units cooperate to perform federated learning using the selected optimal neural network.
• Score includes an index for searching neural networks with excellent learning effects.

• • Potential Applications:**

- Streamlining the process of Neural Architecture Search (NAS). - Enhancing machine learning efficiency and effectiveness. - Enabling practical implementation of machine learning models.

• • Problems Solved:**

- Reducing the time and resources required for NAS. - Improving the accuracy and performance of machine learning models. - Making machine learning more accessible and feasible for various applications.

• • Benefits:**

- Faster and more efficient neural network selection. - Enhanced machine learning capabilities. - Increased practicality and usability of machine learning technology.

• • Commercial Applications:**

Title: "Optimized Neural Network Selection System for Efficient Machine Learning" This technology can be utilized in: - Data analysis and prediction models. - Autonomous systems and robotics. - Personalized recommendation systems.

• • Prior Art:**

Further research can be conducted in the field of federated learning and neural network optimization to explore related prior art.

• • Frequently Updated Research:**

Stay updated on advancements in federated learning, NAS, and machine learning optimization techniques for the latest developments in the field.

• • Questions about Neural Architecture Search (NAS):**

1. How does this learning system improve the efficiency of NAS?

  - The learning system reduces search time by aggregating neural networks based on calculated scores, leading to faster and more effective model selection.

2. What are the key advantages of using federated learning in this context?

  - Federated learning allows for collaborative model training across multiple processing apparatuses, leveraging the selected optimal neural network as a global model for improved performance.

Original Abstract Submitted

Provided are a learning system and the like that significantly shorten a search time of NAS and enable machine learning in a practical time. The learning system includes a learning server apparatus and n processing apparatuses i. The processing apparatus i includes a score calculation unit that calculates a score swhen local data dis applied to each of A neural networks r. The learning server apparatus includes an aggregation unit that aggregates A neural networks using A×n scores sand selects an optimal neural network. The first federated learning unit and the second federated learning units of the n processing apparatuses i cooperate to perform federated learning using the selected optimal neural network as a first global model. The score sincludes an index with which a neural network having an excellent learning effect can be searched for.