EM Estimation for the Bivariate Mixed Exponential Regression Model

EM Estimation for the Bivariate Mixed Exponential Regression Model

In this paper, we present a new family of bivariate mixed exponential regression models for taking into account the positive correlation between the cost of claims from motor third party liability bodily injury and property damage in a versatile manner.


Multivariate zero-modified hurdle models in insurance

Multivariate zero-modified hurdle models in insurance

This paper focuses on Type I multivariate zero truncation and the first case of multivariate zero inflation by employing the multivariate hurdle model to study the aforementioned multivariate zero modification phenomena.


The multivariate mixed Negative Binomial regression model

The multivariate mixed Negative Binomial regression model

This paper is concerned with introducing a family of multivariate mixed Negative Binomial regression models in the context of a posteriori ratemaking.


image-left Differentiation Privacy and Fairness in Automated Decision Making
Read about how Arijit Das tackles the issue of balancing privacy and fairness in automated decison making while maintaing high model accuracy using the UCI Credit-Card Default dataset

image-left A first order binomial mixed Poisson integer valued autoregressive model with serially dependent innovations
This paper is motivated by the extended Poisson INAR(1), which allows innovations to be serially dependent, develops a new family of binomial-mixed Poisson INAR(1) (BMP INAR(1)) processes by adding a mixed Poisson component to the innovations of the classical Poisson INAR(1) process.

image-left Bivariate Mixed Poisson Regression Models with Varying Dispersion
The main purpose of this article is to present a new class of bivariate mixed Poisson regression models with varying dispersion that offers sufficient flexibility for accommodating overdispersion and accounting for the positive correlation between the number of claims from third-party liability bodily injury and property damage.

image-left Federated Learning: Collaboration with no compromise
Read more about the application of federated learning to predict claims frequency as shared by the Federated Learning WP in the September 2021 Data Science Research Section monthly sharing session.

image-left Application of federated learning in predictive maintenance to predict remaining useful life
Learn more as Zack applied two open-source federated learning packages - FATE and dc_federated - to predict failure of turbofan engines.

image-left NLP and ML in Social Media Opinion Mining
Learn more as John introduces NLP and Machine Learning Techniques in social media opinion mining to uncover insights pertinent to public interest and the insurance industry.

image-left Data Visualisation in Insurance
Learn more as Małgorzata introduces visualisation libraries for python.

image-left Modular Framework of Machine Learning Pipeline (Preview)
In this presentation, John Ng introduced the Actuarial Data Science Control Cycle and a modular framework of streamlining analytics and machine learning workflow in enterprises. (IFoA Data Science Webinar Series; IFoA Asia Conference 2020)

image-left Federated Learning for Privacy-Preserving Data Access
Federated learning is a pioneering privacy-preserving data technology and also a new machine learning model trained on distributed data sets. This paper discusses federated learning as a solution for privacy-preserving data access and distributed machine learning applied to distributed data sets. It also presents a privacy-preserving federated learning infrastructure.

image-left Twitter Sentiment Analysis: What does Social Media tell us about coronavirus concerns in the UK?
In this work, we consider the problem of classifying sentiment of UK Twitter messages on COVID-19 using Natural Language Processing (NLP) and supervised Machine Learning techniques.

image-left A Guide for Ethical Data Science (RSS and IFoA)
The Royal Statistical Society and the Institute and Faculty of Actuaries have worked together to jointly produce an ethical framework for their members and practitioners working in the field of data science. Created with practitioners in mind, this guide seeks to provide practical support to members on ethical practice. Structured around our five core ethical themes, the guide provides examples of common ethical challenges in the field and how they could be applied.

Other resources

(*) Highly Recommmended

Data Science in Actuarial Institutes and Societies

  1. Institute and Faculty of Actuaries (IFoA) and Southampton Data Science Academy - Certificate in Data Science
  2. IFoA – Data Science and Lifelong Learning
  3. IFoA – General Insurance Machine Learning in Reserving working party
  4. Swiss Association of Actuaries – Actuarial Data Science tutorials and courses
  5. iCAS – Predictive Analytics and Data Science
  6. Society of Actuaries – Data Analytics Resources
  7. Actuaries Institute of Australia
  8. Singapore Actuarial Society – Data Analytics
  9. International Actuarial Association – Big Data Task Force
  10. Deep Learning with Actuarial Applications in R

Actuarial Publications

  1. The Actuary magazine IFoA – Data Science topics
  2. The Actuary magazine SOA
  3. Annals of Actuarial Science

Books and Publications

  1. (*) An Introduction to Statistical Learning
  2. (*) The Elements of Statistical Learning


  1. (*) Andrew Ng’s Machine Learning Course
  2. Coursera
  3. EdX
  4. Udacity
  5. Udemy
  6. Big Data University


  1. Kaggle
  2. KDnuggets
  3. Towards data science

Data Sources

  1. The Human Mortality Database
  2. Mortality Data Directory, by the Mortality Research Steering Committee (MRSC) and IFoA
  3. Office for National Statistics
  4. CASdatasets package: a collection of insurance datasets
  5. Amazon: Open Data on AWS