Autoencoder Framework for Clinical Disease Prediction on Three Benchmark Datasets

Vijay U. Rathod's study proposes a unified deep learning framework for disease prediction, evaluated on three clinical datasets. The framework, last updated in May 2026, integrates multiple architectures like MLP, CNN, and FT-Transformer with autoencoders. It was tested on the UCI Heart Disease, PIMA Indians Diabetes, and Parkinson's disease voice datasets.

Use Cases

• Benchmarking deep learning architectures for clinical prediction based on the described framework integrating MLP, CNN, and FT-Transformer.
• Developing dataset-aware model selection strategies based on characteristics like feature dimensionality and sample size mentioned in the description.
• Implementing robust evaluation pipelines for clinical AI based on the described fold-safe feature selection and nested cross-validation.
• Building ensemble models for disease classification based on the described FT-Transformer + autoencoder and CNN + Autoencoder architectures.

Strengths

• Framework evaluated on three distinct benchmark datasets with specified sample sizes: UCI Heart Disease (303 samples), PIMA Indians Diabetes (768 samples), and Parkinson's disease (195 recordings).
• Reports specific model performance metrics, such as an AUC of 0.8980 for heart disease prediction and 0.8451 for diabetes classification.
• Includes a detailed methodology with robust preprocessing, fold-safe feature selection, and nested cross-validation as described.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count for the provided configuration data is unknown, which may limit suitability assessment.
• The dataset is very small at 5.5 KB, indicating it likely contains configuration or summary data rather than the primary clinical records.

Provenance

Source

Vijay U. Rathod via figshare

Collection Method

Likely contains configuration parameters or results from the described deep learning framework study.

Freshness

Last updated 2026-05-08 17:34:24; freshness should be verified.