Does scikit-survival work with scikit-learn?

Yes, scikit-survival is built on top of scikit-learn and integrates seamlessly with its API, allowing you to use standard pipelines, GridSearch, and cross-validation.

What is survival analysis in Python?

Survival analysis is a statistical approach used to predict the time until an event occurs. In Python, scikit-survival provides tools to handle 'censored' data where the event hasn't happened yet for some subjects.

Can I use scikit-survival for business churn?

Absolutely. It is highly effective for churn analysis because it doesn't just predict 'if' a customer will leave, but 'when' they are likely to leave, even if they are currently active.

What models can I build with this Claude skill?

You can implement Cox Proportional Hazards models, Random Survival Forests, Gradient Boosting Survival Analysis, and Survival Support Vector Machines (SVM).

How do I evaluate survival models?

This skill guides you in using specific metrics like the Concordance Index (C-index), Brier Score, and time-dependent AUC, which are necessary because standard accuracy metrics don't work with censored data.

scikit-survival Analysis

Name: scikit-survival Analysis
Author: lifangda

bylifangda

•

データサイエンスとML

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

概要

scikit-survival is a specialized extension for scikit-learn that enables advanced time-to-event analysis for datasets with censored observations. This skill empowers Claude to build, train, and evaluate survival models ranging from traditional Cox Proportional Hazards to complex ensemble methods like Random Survival Forests and Gradient Boosting. It provides structured workflows for data preparation, model selection, performance evaluation using concordance indices, and handling competing risks, making it an essential tool for clinical research, churn prediction, and reliability engineering.

主な機能

Specialized survival modeling including Cox, Random Survival Forests, and Survival SVMs
Robust handling of right, left, and interval-censored data types
16 GitHub stars
Advanced evaluation metrics like Harrell's and Uno’s C-index and time-dependent AUC
Integrated workflows for competing risks and non-parametric estimation (Kaplan-Meier)
Full compatibility with scikit-learn pipelines and cross-validation tools

ユースケース

Clinical trial analysis to predict patient survival rates and medical risk factors
Predictive maintenance for industrial equipment to estimate expected time-to-failure
Customer churn prediction by modeling the time until subscription cancellation

概要

主な機能

Specialized survival modeling including Cox, Random Survival Forests, and Survival SVMs
Robust handling of right, left, and interval-censored data types
16 GitHub stars
Advanced evaluation metrics like Harrell's and Uno’s C-index and time-dependent AUC
Integrated workflows for competing risks and non-parametric estimation (Kaplan-Meier)
Full compatibility with scikit-learn pipelines and cross-validation tools

ユースケース

Clinical trial analysis to predict patient survival rates and medical risk factors
Predictive maintenance for industrial equipment to estimate expected time-to-failure
Customer churn prediction by modeling the time until subscription cancellation