Micron technology, inc. (20240289634). TRUST BASED FEDERATED LEARNING simplified abstract

TRUST BASED FEDERATED LEARNING

Organization Name

micron technology, inc.

Inventor(s)

Shashank Bangalore Lakshman of Folsom CA (US)

Pavana Prakash of Houston TX (US)

Poorna Kale of Folsom CA (US)

TRUST BASED FEDERATED LEARNING - A simplified explanation of the abstract

This abstract first appeared for US patent application 20240289634 titled 'TRUST BASED FEDERATED LEARNING

Simplified Explanation

The patent application describes a system where a global model is shared with local devices based on trust signals. The system updates the global model based on feedback from the local devices.

• The host system communicates a global model and loss value to a local device.
• The local device executes a local version of the global model on a local test dataset.
• The host system receives the local loss value and updates the global model if the local loss value is better than the global loss value.

Key Features and Innovation

• Trust-based sharing of global model with local devices.
• Updating the global model based on feedback from local devices.
• Analyzing local loss values to improve the global model.

Potential Applications

• Collaborative machine learning systems.
• Privacy-preserving model training.
• Distributed learning in IoT devices.

Problems Solved

• Efficiently updating global models with local feedback.
• Ensuring privacy and security in federated learning systems.
• Improving model accuracy through localized training.

Benefits

• Enhanced model performance through localized feedback.
• Increased privacy and security in machine learning systems.
• Scalable and efficient distributed learning processes.

Commercial Applications

Title: Trust-Based Federated Learning Systems This technology can be applied in industries such as healthcare for collaborative model training while maintaining data privacy. It can also be used in financial services for secure distributed learning processes.

Prior Art

There may be prior research on federated learning systems and collaborative model training that could provide insights into similar technologies.

Frequently Updated Research

Stay updated on advancements in federated learning systems, collaborative machine learning, and privacy-preserving model training to enhance the efficiency and security of distributed learning processes.

Questions about Federated Learning Systems

How does trust-based sharing of global models improve machine learning processes?

Trust-based sharing ensures that only reliable local devices contribute to the global model, enhancing its accuracy and performance.

What are the potential challenges in implementing federated learning systems in real-world applications?

Challenges may include ensuring data privacy, managing communication between devices, and optimizing model updates across distributed systems.

Original Abstract Submitted

apparatuses and methods related to federated learning are described. a host system can, responsive to a valid trust signal from a first local device, communicate a global model and a global loss value to the local device. the host system can receive a local loss value from the local device. the local loss value can be based on execution of a local version of the global model, generated by the local device, on a local test dataset by the local device. the host system can analyze the local loss value based on quantities of training samples and test samples. responsive to the local loss value being more preferred than the global loss value, the host system can receive the local version of the global model from the local device, update the global model, and communicate the updated global model to the local device and to another local device.