Abstract:As the field of data analysis grows rapidly due to the large amounts of data being generated, effective data classification has become increasingly important. This paper introduces the RUle Mutation Classifier (RUMC), which represents a significant improvement over the Rule Aggregation ClassifiER (RACER). RUMC uses innovative rule mutation techniques based on evolutionary methods to improve classification accuracy. In tests with forty datasets from OpenML and the UCI Machine Learning Repository, RUMC consistently outperformed twenty other well-known classifiers, demonstrating its ability to uncover valuable insights from complex data.

What problem does this paper attempt to address?

### What problems does this paper attempt to solve? This paper aims to solve the limitations of existing algorithms in data classification when dealing with high - dimensional and low - sample - size data. In particular, the initial rules generated by the Rule Aggregation Classifier (RACER) are too specific, resulting in poor generalization ability on new data. Specifically: 1. **Rules are too specific**: The initial rules generated by RACER are usually too specific, which makes them perform poorly when dealing with new data, especially on data sets with small sample sizes. 2. **Difficulties in processing high - dimensional data**: When the number of features far exceeds the number of samples, many algorithms (including RACER) are difficult to accurately estimate parameters, leading to performance degradation. 3. **Improving classification accuracy**: Existing classification algorithms often fail to achieve ideal classification accuracy when dealing with complex data, especially in diverse data environments. To solve these problems, the author proposes the RUle Mutation Classifier (RUMC), which is a rule - mutation classifier based on evolutionary methods. RUMC iteratively optimizes the initial rule set by introducing the rule - mutation technique, thereby improving classification accuracy and outperforming other well - known classifiers on a variety of data sets. ### Main improvement points of RUMC - **Rule mutation**: RUMC uses the mutation mechanism in evolutionary algorithms to optimize the initial rules to generate more generalizable rules. - **Wide applicability**: RUMC not only performs well on high - dimensional and low - sample - size data, but also can maintain high classification accuracy on large - scale data sets. - **Efficient rule combination**: By performing a logical "OR" operation on the rules, RUMC can create more widely applicable rules, thereby improving classification performance. Through these improvements, RUMC significantly outperformed 20 other classifiers in tests on 40 public data sets, demonstrating its potential and advantages in the field of data classification. ### Summary The core problem of this paper is to improve the classification accuracy of rule - based classifiers in complex data environments, especially in response to the challenges of high - dimensional and low - sample - size data. RUMC successfully solves the limitations of existing algorithms by introducing the rule - mutation technique and the idea of evolutionary algorithms, providing higher classification accuracy and better generalization ability.

RUMC: A Rule-based Classifier Inspired by Evolutionary Methods

ROPAC: Rule OPtimized Aggregation Classifier

Machine Learning and Evolutionary Computing for GUI-based Regression Testing

Improved Runge Kutta Optimization Using Compound Mutation Strategy in Reinforcement Learning Decision Making for Feature Selection

Evolutionary Optimization of Model Merging Recipes

Adaptive Memetic Computing for Evolutionary Multiobjective Optimization

Mutation Testing framework for Machine Learning

MuRS: Mutant Ranking and Suppression using Identifier Templates

Searching for Interpretable Rules for Disease Mutations: a Simulated Annealing Bump Hunting Strategy

A Bandit Approach with Evolutionary Operators for Model Selection

EVOTER: Evolution of Transparent Explainable Rule-sets

Optimized RMDL with Transfer Learning for Sentiment Classification in the MapReduce Framework

GPU-Accelerated Rule Evaluation and Evolution

Mutation-driven Test Case Generation Using Short-lived Concurrent Mutants -- First Results

A Memetic Algorithm Using Emperor Penguin and Social Engineering Optimization for Medical Data Classification

Guided Evolution with Binary Discriminators for ML Program Search

Towards Mutation Testing of Reinforcement Learning Systems

A cooperative coevolution-based pittsburgh learning classifier system embedded with memetic feature selection

MULTICR: Predicting Merged and Abandoned Code Changes in Modern Code Review Using Multi-Objective Search

Deep-GenMut: Automated genetic mutation classification in oncology: A deep learning comparative study

MCMC implementation of the optimal Bayesian classifier for non-Gaussian models: model-based RNA-Seq classification