Microsoft technology licensing, llc (20240338532). DISCOVERING AND APPLYING DESCRIPTIVE LABELS TO UNSTRUCTURED DATA simplified abstract

DISCOVERING AND APPLYING DESCRIPTIVE LABELS TO UNSTRUCTURED DATA

Organization Name

microsoft technology licensing, llc

Inventor(s)

Wolfgang Martin Pauli of Seattle WA (US)

Robert McArn Horton of Corte Madera CA (US)

DISCOVERING AND APPLYING DESCRIPTIVE LABELS TO UNSTRUCTURED DATA - A simplified explanation of the abstract

This abstract first appeared for US patent application 20240338532 titled 'DISCOVERING AND APPLYING DESCRIPTIVE LABELS TO UNSTRUCTURED DATA

• • Simplified Explanation:**

The patent application discusses a method for training machine learning models using a combination of human-annotated samples and soft labels generated by a large language machine learning model.

• • Key Features and Innovation:**
• Selecting training samples from a dataset
• Generating soft labels using a large language machine learning model
• Training a student model with the training samples
• Evaluating the student model's performance based on human-annotated samples
• Selecting additional training samples using a teacher model
• Retraining the student model with the new training samples

• • Potential Applications:**

This technology can be applied in various fields such as natural language processing, computer vision, and speech recognition to improve the accuracy and efficiency of machine learning models.

• • Problems Solved:**

This technology addresses the challenge of training machine learning models with limited human-annotated data by leveraging soft labels generated by a large language machine learning model.

• • Benefits:**
• Improved model performance with limited human-annotated data
• Enhanced efficiency in training machine learning models
• Increased accuracy and generalization of models

• • Commercial Applications:**

Potential commercial applications include automated content moderation, sentiment analysis, and personalized recommendations in e-commerce platforms. This technology can also be used in autonomous vehicles for object detection and recognition.

• • Questions about the Technology:**

1. How does this technology improve the training process of machine learning models? 2. What are the advantages of using soft labels generated by a large language machine learning model in training?

• • Frequently Updated Research:**

Researchers are constantly exploring new methods to enhance the training process of machine learning models, including the use of semi-supervised learning techniques and advanced data augmentation strategies. Stay updated on recent developments in this field to leverage the latest advancements in machine learning training methodologies.

Original Abstract Submitted

example solutions for training machine learning models include: selecting a plurality of training samples from a dataset; generating soft labels for the training samples using a large language machine learning model (llm); training a student model using the plurality of training samples; evaluating a performance metric of the student model based on a plurality of human-annotated samples; selecting one or more additional training samples from the dataset using a teacher model; generating soft labels for the one or more additional training samples using the llm; and retraining the student model using at least the plurality of training samples and the one or more additional training samples.