Evaluation of Feature Selection Techniques for Predicting Parkinson's Disease using Machine Learning Models

Created by W.Langdon from gp-bibliography.bib Revision:1.8415

@InProceedings{Akilandeswari:2025:ICEARS,

• author = "A. Akilandeswari",
• title = "Evaluation of Feature Selection Techniques for Predicting {Parkinson's} Disease using Machine Learning Models",
• booktitle = "2025 International Conference on Electronics and Renewable Systems (ICEARS)",
• year = "2025",
• pages = "1431--1435",
• month = feb,
• keywords = "genetic algorithms, genetic programming, Logistic regression, Renewable energy sources, Accuracy, Parkinson's disease, Pipelines, Machine learning, Predictive models, Feature extraction, Random forests, Optimisation, Parkinson's Disease Prediction, Random Forest, Boruta, Recursive Feature Elimination (RFE), TPOT",
• DOI = "doi:10.1109/ICEARS64219.2025.10940164",
• abstract = "Parkinson's disease is one of the major progressive neurodegenerative disorders that majorly affect both motor and non-motor functions. The significance of this study lies in spotlighting the importance of early, accurate detection of Parkinson's disease in the betterment of patient outcomes as well as management. This paper sets out to make a case discussion about the use of machine learning models in predictive Parkinson's disease through feature selection. We apply the following methods to machine learning models, including Random Forest and Logistic Regression, in order to compare approaches, namely Boruta, Recursive Feature Elimination with Cross-Validation (RFECV), TPOT-Tree-based Pipeline Optimisation Tool, and Variance Threshold. The Random Forest model with Variance Threshold feature selection showed a high accuracy of 95percent. Our experiments also present the fact that feature selection is relevant to improve the performance of the model: indeed, different techniques show different levels of prediction accuracy. ConclusionThis paper has proven it possible to improve the prediction of disease to a considerable extent by the right selection of the feature selection technique and machine learning model.",
• notes = "Also known as \cite{10940164}",

}

Genetic Programming entries for A Akilandeswari

Citations