Course structure
Session 1
- How to create value.
- Stakeholder analysis.
- What is the use case and value story?
- Introduction to analytics method.
- How to use the Analytics Use Case Canvas.
- How to be agile when working with analytics.
- Introduction to insurance methods.
- Read data from different DB (data lake and data mart).
- Data exploration and diagnostic.
- Data partitioning and honest assessment procedures (Bootstrap and k-fold cross-validation).
- Data quality and data transformation.
- Profit loss matrix.
Session 2
- Stakeholder analysis and interviewing techniques.
- Prepare, explore, and visualize data.
- Predictive analytics.
- Formulate your own analytics use case.
- Data-analytics personas.
- Feature extraction and variable selection methods.
- Claims, churn, and fraud model prediction.
- Triaging claims.
- Pricing and risk selection (pricing improvement).
- Customer segmentation.
Session 3
- Open source and cloud.
- How to implement and take action on analytics.
- Evaluating and monitoring analytics.
- Best practice for deployment analytics.
- Decision process and need analysis.
- Identify themes in claims and transcriptions: text analysis (collect documents, corpus parsing, topic modeling, word embedding, and sentiment analysis).
- Model deployment.
Session 4
- Advanced data preparation and exploration.
- Advanced predictive modeling.
- From informer to influencer.
- Transition model.
- Change model.
- Policy recommendation engine.
- Personalized marketing.
- Customer lifetime value.
- Real-time monitoring of pricing models.
Session 5
- Presentation technique.
- Lifelong learning.
- Machine learning and computer vision.
- Natural language analytics and optimization.
- What are neural networks and deep learning?
- Ethics, bias, and explainability in AI.
- AI next generation.
- Computer vision in insurance.
