scikit-survival-analysis

What problem does it solve?

This Skill enables sophisticated analysis of time-to-event data, crucial for understanding patient survival, equipment reliability, or customer churn, especially when dealing with censored observations.

Core Features & Use Cases

• Model Fitting: Supports Cox proportional hazards, Random Survival Forests, Gradient Boosting, and SVMs for censored data.
• Evaluation: Provides censoring-aware metrics like C-index and Integrated Brier Score.
• Data Handling: Includes utilities for preparing survival data structures and handling competing risks.
• Use Case: Predict patient survival probability based on clinical features, accounting for patients who are still alive at the end of the study (censored).

Quick Start

Use the scikit-survival-analysis skill to fit a Random Survival Forest model to the provided training data and evaluate its performance on the test set.