Adaptive Machine Learning Control Using Input Risk Stress
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Machine learning systems face challenges in adapting to changes in the external environment, leading to suboptimal actions and lower rewards due to external environment changes or past learning errors.
Innovation Solution
An information processing system that includes a learning section, an input information assessment section, and a first parameter calculation section to assess risk and calculate a stress parameter, adjusting learning efficiency based on this assessment to adapt to changing environments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the system performs machine learning to optimize action selection, then learning capability is improved, but adaptability to external environment changes deteriorates
Solution Approach 1:
The patent implements dynamic adjustment of learning efficiency based on the stress parameter. The learning section changes its learning efficiency according to the first parameter (stress) calculated from input information assessment. This allows the system to adapt its learning behavior dynamically - increasing learning efficiency when stress is high (indicating environmental changes) and maintaining normal efficiency when stress is low, thereby resolving the contradiction between adaptability and learning reliability
Solution Approach 2:
The system changes the parameter of learning efficiency based on the stress parameter. By calculating the stress parameter from assessed input information and using it to modulate learning efficiency, the system transforms a static learning process into a dynamic one that responds to environmental changes, improving adaptability while maintaining reliability through controlled parameter adjustment
2Adaptability or versatility
If the system increases learning efficiency to adapt quickly to changes, then adaptability is improved, but learning stability deteriorates
Solution Approach 1:
The learning efficiency is dynamically adjusted based on the stress parameter rather than being fixed. When the input information assessment section detects high stress (indicating environmental changes), the learning section increases learning efficiency to adapt quickly. When stress is low, learning efficiency returns to normal levels, maintaining stability. This dynamic adjustment mechanism resolves the contradiction between adaptability and stability
3Adaptability or versatility
If the system uses fixed learning parameters, then learning stability is maintained, but adaptability to environment changes deteriorates
Solution Approach 1:
The system performs self-adjustment of learning efficiency based on its own stress parameter calculation. The learning section automatically changes its learning efficiency according to the first parameter calculated from its own input information assessment, without requiring external control or complex configuration. This self-service mechanism improves environmental adaptability while keeping the control complexity manageable
Data Source
AI summary
The present disclosure relates to an information processing system, an information processing method, and an information processing device that make it possible to perform learning adaptively to changes in an external environment and circumstances. A learning section learns results of action selection made by a system in response to input information. An input information assessment section assesses a risk of the input information to the system. A first parameter calculation section calculates a first parameter representing stress on the system according to an assessed value of the input information. The learning section changes learning efficiency according to the first parameter. The technology according to the present disclosure is applicable, for example, to an information processing system that performs machine learning.