Adaptive Threshold Causal Search in Attribute Correlation Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for causal relationship search in numerical data are inefficient due to variability in correlation detection based on binarization threshold value settings, leading to inconsistent results.
Innovation Solution
An information processing program that generates binary data from original feature amounts, enumerates frequently appearing saturated item sets, computes correlations, and selects conditions with adaptive threshold updates to ensure accurate correlation detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If binarization with fixed threshold values is used to extract correlation conditions, then the processing can be simplified, but the measurement precision of correlation detection deteriorates due to variability based on threshold settings
Solution Approach 1:
The patent changes the parameter of threshold values from fixed to adaptive. The threshold values are dynamically adjusted based on the distribution characteristics of the actual data, allowing the binarization process to adapt to different datasets while maintaining both simplicity and precision in correlation detection
Solution Approach 2:
The patent implements a feedback mechanism where the system evaluates the distribution of actual data and uses this information to adjust the threshold values. This feedback loop ensures that the threshold settings are optimized for the specific dataset being analyzed, improving correlation detection accuracy without increasing processing complexity
2Reliability
If all condition candidates are searched for causal relationships, then the comprehensiveness of causal discovery is improved, but the productivity deteriorates due to unrealistic computation amounts
Solution Approach 1:
The patent extracts only the most relevant condition candidates by using adaptive binarization to identify significant correlations. Instead of processing all possible conditions, the system extracts a focused subset of high-probability causal relationships, dramatically reducing computation while maintaining comprehensiveness
Solution Approach 2:
The patent segments the large search space of all condition candidates into manageable subsets based on adaptive threshold binarization. By dividing the problem into segments processed at different threshold levels, the system efficiently explores causal relationships without being overwhelmed by the full computation burden
Data Source
AI summary
A non-transitory computer-readable recording medium stores an information processing program for causing a computer to execute processing including: generating, from first data in which values of a plurality of attributes included in each sample are accumulated for each sample, second data obtained by binarizing, for each sample, the values of the plurality of attributes included in each sample based on an attribute condition set in advance; enumerating, by using the second data, sets of attribute conditions in which all sample sets indicate true values; computing, for each set of attribute conditions, a correlation between the plurality of attributes in the first data in a sample set associated with each set of attribute conditions; and selecting a set of attribute conditions determined to have a correlation as a condition to be causally searched.