Optimizing Alzheimer's Disease Diagnostic Accuracy: a Comparative Study of T-SNE, Autoencoders, and SVM on ADNI Dataset

View at publisher Summarise with VIRAT
Authors: Shruti P, Rekha R, Prasana B, Ragul Rishwanth P, Pranovesh K J, Jayavasanth D
Cited by: 0
References: 25

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

This study compares the effectiveness of various dimensionality reduction (DR) techniques-including Principal Component Analysis (PCA), Uniform Manifold Approximation and Projection (UMAP), Multidimensional Scaling (MDS), and t-Distributed Stochastic Neighbor Embedding (t-SNE)-combined with a Support Vector Machine (SVM) classifier for a diagnostic task. The results showed that applying DR significantly enhanced performance over the baseline SVM (Accuracy: 0.68). Specifically, the fusion of t-SNE and an Autoencoder achieved the highest accuracy at 94 %, outperforming all other configurations. This demonstrates that leveraging advanced non-linear DR and feature learning methods provides a superior and highly accurate framework for classification.

Fields of Study & Subjects

Dementia and Cognitive Impairment Research Artificial Intelligence in Healthcare Face and Expression Recognition

Publication details

Journal / source
International Conference on Intelligent Systems for Communication, IoT and Security (ICISCoIS)
Pages
1–7
Publication date
8 January 2026
Language
EN