Qualcomm incorporated (20240119360). ADAPTING MACHINE LEARNING MODELS FOR DOMAIN-SHIFTED DATA simplified abstract
Contents
ADAPTING MACHINE LEARNING MODELS FOR DOMAIN-SHIFTED DATA
Organization Name
Inventor(s)
ADAPTING MACHINE LEARNING MODELS FOR DOMAIN-SHIFTED DATA - A simplified explanation of the abstract
This abstract first appeared for US patent application 20240119360 titled 'ADAPTING MACHINE LEARNING MODELS FOR DOMAIN-SHIFTED DATA
Simplified Explanation
The present disclosure provides techniques for adapting a machine learning model for inferencing against a target data set in a shifted domain from a source data set used to train the model.
- Identifying domain-sensitive layers in a machine learning model based on differences between outputs generated for inputs in the source domain and inputs in the shifted domain.
- Updating normalizing values for each domain-sensitive layer based on a mixing factor, fixed normalizing values for data in the source domain, and calculated normalizing values for data in the shifted domain.
- Applying the updated normalizing values to each domain-sensitive layer in the machine learning model.
- Potential Applications:**
- Adapting machine learning models for different domains - Improving the performance of machine learning models in shifted domains
- Problems Solved:**
- Addressing domain shift in machine learning models - Enhancing the generalization capabilities of machine learning models
- Benefits:**
- Increased accuracy in inferencing against target data sets - Improved adaptability of machine learning models to new domains
- Potential Commercial Applications:**
- Customizing machine learning models for specific industries - Enhancing the performance of predictive analytics tools
- Possible Prior Art:**
One possible prior art could be techniques for domain adaptation in machine learning models using transfer learning approaches.
- Unanswered Questions:**
- 1. How does the mixing factor impact the updating of normalizing values in domain-sensitive layers?**
The specific relationship between the mixing factor and the updating process of normalizing values could provide more insights into the adaptation process.
- 2. Are there any limitations to the effectiveness of this technique in highly complex or specialized domains?**
Understanding the potential constraints or challenges faced by this adaptation technique in certain domains could help in assessing its applicability in various scenarios.
Original Abstract Submitted
certain aspects of the present disclosure provide techniques and apparatuses for adapting a machine learning model for inferencing against a target data set in a shifted domain from a source data set used to train the machine learning model. an example method generally includes identifying one or more domain-sensitive layers in a machine learning model based on differences between outputs generated by one or more layers in the machine learning model for inputs in a source domain and inputs in a shifted domain. normalizing values are updated for each respective domain-sensitive layer of the one or more domain-sensitive layers based on a mixing factor, fixed normalizing values for data in the source domain, and calculated normalizing values for data in the shifted domain. the updated normalizing values are applied to each respective domain-sensitive layer of the one or more domain-sensitive layers in the machine learning model.
