Depression Prediction in Haemodialysis Patients Using PCA and Machine Learning

1.3 MB of research data and code for predictive modelling of depression in patients undergoing chronic haemodialysis. The study by FRIH Hacène, last updated in May 2026, combined Principal Component Analysis with Logistic Regression, AdaBoost, XGBoost, and CatBoost to identify predictive patterns. AdaBoost and XGBoost demonstrated the strongest predictive performance during independent external validation.

Use Cases

• Predicting depression vulnerability based on multidimensional demographic, clinical, and psychosocial factors mentioned in the description
• Comparing the performance of machine learning models like AdaBoost and XGBoost for clinical classification tasks
• Developing early psychological risk identification tools for personalized mental health assessment in nephrology

Strengths

• Dataset is 1.3 MB in size, indicating a focused and manageable scope for analysis.
• The study methodology involved independent external validation, which suggests a rigorous evaluation approach.
• The dataset is associated with multiple file formats (IPYNB, DOCX, TXT, XLSX), potentially offering code, documentation, and raw data.

Limitations

• Row count is unknown, which may limit suitability assessment for large-scale model training.
• Column-level documentation is absent; field semantics must be inferred after download.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

figshare

Collection Method

Likely collected from clinical studies of haemodialysis patients, as inferred from the description.

Freshness

Last updated 2026-05-17 21:15:52; freshness should be verified.