Exploring Computational Models for Parkinson’s Disease Diagnosis: Unveiling Insights with LIME and SHAP Explainability Techniques

Parkinson’s disease (PD) is a degenerative neurological disorder affecting a large portion of the world’s population, mostly the elderly. A patient has to be diagnosed with Parkinson’s disease (PD) as soon as feasible to receive appropriate management and intervention. This is because early diagnosis opens up treatment choices that can significantly improve patient’s quality of life. To develop advanced diagnostic models for Parkinson’s disease, there has been a notable rise in research at the intersection of computer science and medicine. Preprocessing data, choosing pertinent features, and training models-such as Decision Trees, SVM, K-Nearest Neighbors, Logistic Regression, and other boosting algorithms-are all part of the research. To diagnose Parkinson’s disease, the comparison analysis evaluates model performance indicators and emphasizes interpretability using LIME and SHAP approaches. The usefulness, benefits, and drawbacks of the many computer models utilized in Parkinson’s disease diagnosis are thoroughly examined in this article.