Information Processing Apparatus Using Recent Data for Discrete Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing black box optimization methods for discrete parameters are inefficient and lack accuracy, particularly in dynamically changing environments, due to reliance on outdated history data and suboptimal surrogate model approximations.
Innovation Solution
An information processing apparatus that executes estimation, optimization, sampling, and addition processing to dynamically update a history data set by deleting older data, ensuring the surrogate model is based on recent data for improved accuracy and responsiveness.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the surrogate model is updated using all historical data, then the model approximation accuracy is improved, but the processing time and computational complexity increase
Solution Approach 1:
The patent extracts only the most recent data points from the historical data set that are necessary for maintaining accurate surrogate model approximation. By selecting a subset of recent data rather than using all historical data, the system reduces computational complexity and processing time while maintaining model accuracy for dynamically changing functions.
Solution Approach 2:
The patent implements a dynamic data selection mechanism that adapts to changes in the black box function. The system dynamically determines which historical data points to retain based on the current state and characteristics of the function being optimized, allowing the surrogate model to remain accurate without processing the entire historical data set.
2Measurement precision
If the history data set retains all previous data, then the surrogate model approximation is more accurate, but the memory usage and data processing load increase
Solution Approach 1:
The patent extracts and retains only the essential recent data points from the historical data set. By filtering out redundant or less relevant historical data, the system maintains sufficient information for accurate surrogate model approximation while significantly reducing the total data volume that must be stored and processed.
3Measurement precision
If sampling is performed frequently to update the surrogate model, then the optimization accuracy is improved, but the number of black box function evaluations increases
Solution Approach 1:
The patent applies partial action by performing surrogate model updates only when necessary rather than at every optimization step. The system determines the appropriate frequency of sampling and model updates based on the current optimization state and function characteristics, reducing the total number of black box function evaluations while maintaining optimization accuracy.
Data Source
AI summary
According to one embodiment, an information processing apparatus includes a hardware processor configured to repeatedly execute estimation processing, optimization processing, sampling processing, and addition processing. The hardware processor is configured to, in the estimation processing, estimate a surrogate model based on a history data set. The hardware processor is configured to, in the optimization processing, calculate, as N discrete parameters, a solution to an optimization problem of minimizing or maximizing the surrogate model. The hardware processor is configured to, in the sampling processing, sample an output value. The hardware processor is configured to, in the addition processing, add history data including the N discrete parameters and the output value to the history data set. The hardware processor is configured to further execute deletion processing of deleting at least a part of one or more pieces of history data from the history data set.