Ancestry Count Tracking in Genetic Algorithm Fitness Evaluation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Genetic algorithms used in data mining often face challenges such as premature convergence to local optima and the inability to accurately assess the fitness of individuals due to incomplete testing, leading to optimization for 'lucky' rather than truly fit individuals.
Innovation Solution
Implementing an experience-layered elitist pool structure where individuals are stratified by their testing experience levels, with a competition module that handicaps or alters mutation rates based on ancestry count, and sets minimum fitness criteria to ensure more accurate fitness evaluations and reduce the impact of lucky individuals.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If genetic algorithms are used to mine large data sets, then useful patterns can be extracted, but premature convergence to local optima occurs and fitness evaluation becomes inaccurate
Solution Approach 1:
The patent segments the population into experience layers based on the number of generations an individual has survived. This segmentation allows individuals with different testing experiences to be evaluated and competed separately, preventing premature convergence while maintaining accurate fitness evaluation. Each layer represents a distinct segment of the evolutionary process, enabling more reliable selection decisions.
Solution Approach 2:
The patent applies preliminary action by handicapping individuals based on their ancestry count before fitness evaluation. This preemptive adjustment compensates for the incomplete testing that occurs when individuals are evaluated after only a few generations, ensuring that fitness comparisons are more accurate from the outset and preventing lucky individuals from dominating early evaluations.
2Productivity
If individuals are tested on limited training data, then computational resources are conserved, but fitness estimates become inaccurate due to incomplete testing
Solution Approach 1:
The patent implements feedback by continuously tracking the ancestry count of each individual and using this information to adjust fitness evaluations. The system feeds back the evolutionary history information into the selection process, allowing fitness estimates to be corrected based on how many generations an individual has survived, thereby improving accuracy without requiring exhaustive testing of every individual.
Solution Approach 2:
The patent changes the fitness evaluation parameter by introducing ancestry count as a modifying factor. Instead of using raw fitness scores alone, the system adjusts evaluation parameters based on the number of generations an individual has persisted, allowing accurate fitness estimation with limited testing data by accounting for the individual's evolutionary experience.
3Loss of time
If the population converges to local optima, then computational time is reduced, but the quality of extracted patterns deteriorates
Solution Approach 1:
The patent applies dynamics by making the selection process adaptive based on ancestry count. Rather than using a static selection criterion, the system dynamically adjusts selection probabilities and fitness evaluations based on how many generations each individual has survived. This dynamic approach prevents premature convergence to local optima while maintaining efficient execution by avoiding unnecessary generations.
4Measurement precision
If ancestry count tracking is implemented, then fitness evaluation accuracy improves, but system complexity increases
Solution Approach 1:
The patent achieves universality by making the ancestry count serve multiple functions simultaneously. The same ancestry count data structure is used for handicapping individuals, for layer segmentation, and for adjusting fitness evaluations. This multi-functional use of a single data element improves fitness evaluation accuracy without proportionally increasing system complexity, as one piece of information accomplishes multiple corrective tasks.
Data Source
AI summary
Roughly described, a computer-implemented evolutionary data mining system includes a memory storing a candidate gene database in which each candidate individual has a respective fitness estimate; a gene pool processor which tests individuals from the candidate gene pool on training data and updates the fitness estimate associated with the individuals in dependence upon the tests; and a gene harvesting module for deploying selected individuals from the gene pool, wherein the gene pool processor includes a competition module which selects individuals for discarding in dependence upon their updated fitness estimate. The system maintains the ancestry count for each of the candidate individuals, and may use this information to adjust the competition among the individuals, to adjust the selection of individuals for further procreation, and/or for other purposes.