What is scikit-survival?

It is a Python library built on top of scikit-learn that extends its capabilities to survival analysis, allowing for the modeling of time-to-event data where observations may be censored.

How does it handle model evaluation?

The skill utilizes specialized metrics that account for censoring, such as Harrell’s and Uno’s C-index, as well as time-dependent AUC and integrated Brier scores for both discrimination and calibration.

When should I use this skill instead of standard regression?

Use this skill when your target variable involves a duration until an event occurs and some observations are 'censored,' meaning the event hasn't happened yet for those subjects during the observation period.

Does this skill support high-dimensional data?

Yes, it includes specialized models like CoxnetSurvivalAnalysis which uses Elastic Net regularization to handle datasets where the number of features is large relative to the number of samples.

Can I use scikit-learn Pipelines with this skill?

Absolutely. Scikit-survival is designed to be fully compatible with scikit-learn's ecosystem, including Pipelines, StandardScaler, and GridSearchCV for automated hyperparameter tuning.

Scikit-Survival Analysis

Name: Scikit-Survival Analysis
Author: Zehong-Wang

byZehong-Wang

0•

Data Science & ML

Performs comprehensive survival analysis and time-to-event modeling using the scikit-survival library in Python.

This skill empowers Claude to conduct sophisticated survival analysis, specifically designed for datasets with censored observations where events may not have occurred during the study period. It provides a robust framework for fitting various models—ranging from interpretable Cox proportional hazards to high-performance ensemble methods like Random Survival Forests—while ensuring proper evaluation through specialized metrics like Uno's C-index and Brier scores. Whether you are handling clinical trial data or churn prediction, this skill streamlines the entire workflow from preprocessing and feature selection to model comparison and validation within the scikit-learn ecosystem.

Key Features

01Provides specialized evaluation metrics like Concordance Index (C-index) and time-dependent AUC.

02Supports advanced ensemble methods including Random Survival Forests and Gradient Boosting.

030 GitHub stars

04Handles complex survival scenarios including competing risks and various censoring types.

05Implements Cox proportional hazards and penalized Coxnet models for high-dimensional data.

06Integrates seamlessly with scikit-learn pipelines, cross-validation, and grid search tools.

Use Cases

01Predicting patient survival rates and risk factors in clinical research or healthcare settings.

02Analyzing customer churn and time-to-attrition for subscription-based businesses.

03Predicting equipment failure and maintenance schedules in industrial reliability engineering.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add zehong-wang/kosmos scikit-survival

For use in Claude.ai and ChatGPT