Talks S2E6 (Louise Ferbach): Deep Learning For Survival Analysis | Summary and Q&A

September 17, 2021
Abhishek Thakur
YouTube video player
Talks S2E6 (Louise Ferbach): Deep Learning For Survival Analysis


This presentation provides an overview of deep learning models for survival analysis, including the Cox regression model and the deep survival model, explaining the concepts of survival analysis, censoring, and truncation. It also discusses the application of these models in various fields, such as insurance, customer analytics, and healthcare.

Install to Summarize YouTube Videos and Get Transcripts

Key Insights

  • ❓ Survival analysis is a statistical method for analyzing the duration before the occurrence of an event, such as death or failure.
  • 🏑 Survival analysis has various applications in different fields, including insurance, credit risk, engineering, public policies, and customer analytics.
  • ❓ Censoring and truncation are common challenges in survival analysis due to incomplete or limited data.
  • ❓ Deep learning models like DeepSurv and CoxTime have been developed to improve survival analysis and provide more accurate predictions.
  • 📈 Evaluating survival models can be done through metrics like the concordance index, but additional metrics may be needed to assess prediction accuracy and model performance.


hello everyone and welcome to this new episode of talks today we have with us luis ferbach and she is going to talk about deep learning for survival analysis so luis sorry if i'm not pronouncing your name right but yeah and she's working with score where she is working as an actuarial data scientist so what does an actual data scientist do luis so ... Read More

Questions & Answers

Q: What is the role of an actuarial data scientist in the insurance industry?

An actuarial data scientist specializes in statistical analysis and risk assessment in the insurance sector. They use data to make predictions about mortality rates, assess global trends, and develop pricing models for insurance products.

Q: What are the applications of survival analysis in customer analytics?

Survival analysis is widely used in customer analytics, particularly for predicting churn or customer retention. By analyzing customer lifetimes, companies can estimate how long customers are likely to stay with them and identify factors that influence their decision to churn.

Q: Can survival analysis be applied to electronic health records (EHR) databases?

Yes, survival analysis can be applied to EHR databases. By analyzing the duration of certain medical events or treatments, survival analysis can help estimate patient outcomes, predict disease progression, and assess the effectiveness of interventions.

Q: Is it possible to transform survival analysis into a binary classification problem?

Survival analysis can be transformed into a binary classification problem by predicting whether an event will occur within a given fixed period. This can be achieved by setting a specific time threshold and classifying events that occur within that period as positive and events that don't occur as negative.

Summary & Key Takeaways

  • The speaker, Luis Ferbach, introduces survival analysis and its applications in fields like insurance, credit risk, engineering, and public policies.

  • She explains the statistical framework of survival analysis, including concepts like censoring and truncation, as well as the calculation of hazard rates and survival functions.

  • The presentation focuses on two deep learning models for survival analysis: DeepSurv and CoxTime. These models extend the Cox proportional hazards model and can be used to estimate survival probabilities and rank individuals based on their risk.

Share This Summary 📚

Summarize YouTube Videos and Get Video Transcripts with 1-Click

Download browser extensions on:

Explore More Summaries from Abhishek Thakur 📚

Summarize YouTube Videos and Get Video Transcripts with 1-Click

Download browser extensions on: