Data Science for Insurance Analytics
June 21th to July 6th, 2022
¡Valoramos tu interés!
La inscripción a este programa no está disponble en este momento
Descubre otros programas de tu interés disponibles.
Organize
Dates
June 21th to July 6th, 2022
Schedule
Monday to Friday from 6:00 p.m. to 8:00 p.m. and Saturdays from 9:00 a.m. to 11:00 a.m.
Venue
In-person
Intensity
24 hours
Requirements
- For U.Rosario undergraduate students: passed “Métodos cuantitativos en finanzas”.
- For postgraduate students U.Rosario: passed “Programación y modelación estocástica”.
- For external participants: prior knowledge of probability and statistical inference.
Fees
- $1.600.000 with grade and course attendance certificate or $800.000 only with course attendance certificate
Discounts
- For companies: discount for group registration. Four (4) people enrolled for the price of three (3)
Description
Emphasizes data science techniques and introductory applications to actuarial science, finance, and economics. This course will give an overview of the different data science methods and algorithms that can be employed to discover useful information from real-life datasets, to explain how to build a data science model using computational software packages (R and Python), and to effectively communicate the results in a scientific report.
Topics include identifying the business problem, basic database calculations and manipulations, data and model ethics, data preparation, data visualization, model building processes, advanced predictive analytics models, model selection, refinement, and validation, model explainability. We will cover case studies in different fields in finance and insurance. For instance, auto/home insurance claims, stock and option price prediction, mortality trends, health data analytics, etc.
- Students will identify different statistical learning methods and algorithms that can be employed to discover useful information from datasets, to explain how to build a predictive model, and to communicate the results in a scientific report.
- Students will have a comprehensive understanding of supervised (regression and classification) and unsupervised (clustering) learning problems.
- Students will be able to implement these methods in practice using a statistical computing environment
- Students will understand especially the bias-variance trade-off of machine learning methods, and understand how these properties are achieved using tuning hyperparameters, and further are able to discuss data and model ethics.
12 sessions 100% In-person
General objective
The main objective of this course is to introduce data science techniques and their applications to actuarial science, finance, and economics.
Specific objectives
- Give an overview of several data science methods and algorithms that can be employed to discover useful information from real-life datasets.
- Explain how to build a data science model using computational software packages (R and Python).
- Describe how to effectively communicate data science results in a scientific report.
Professor
Zhiyu (Frank) Quan is an assistant professor of actuarial science at the University of Illinois at Urbana-Champaign.
Zhiyu Quan
Thematic content
Syllabus
- Introduction to data science.
- Insurance analytics problems.
- Business problem definition.
Professor
- Zhiyu Quan.
Date
Tuesday, June 21th
Schedule
6:00 p.m. to 8:00 p.m
Workload
2 hours
Syllabus
- Data collection practices and assumptions.
- Basic database calculations.
- Extract, transform, and load (ETL) operations.
- Data quality.
- Data consistency.
- Relational databases.
Professor
- Zhiyu Quan.
Date
Wednesday, June 22th
Schedule
6:00 p.m. to 8:00 p.m
Workload
2 hours
Syllabus
- Data types and exploration.
- Data issues and cleaning.
- Data visualization.
Professor
- Zhiyu Quan.
Date
Thursday, June 23th
Schedule
6:00 p.m. to 8:00 p.m
Workload
2 hours
Syllabus
- Standards of actuarial practice.
- Regulations.
- Analytical bias.
Professor
- Zhiyu Quan.
Date
Friday, June 24th
Schedule
6:00 p.m. to 8:00 p.m
Workload
2 hours
Syllabus
- Linear model.
- Generalized linear models.
- Regularization.
- Linear mixed models.
Professor
- Zhiyu Quan.
Date
Saturday, June 25th
Schedule
9:00 a.m. to 11:00 a.m
Workload
2 hours
Syllabus
- Tree-based models.
Professor
- Zhiyu Quan.
Date
Tuesday, June 28th
Schedule
6:00 p.m. to 8:00 p.m
Workload
2 hours
Syllabus
- Neural networks.
Professor
- Zhiyu Quan.
Date
Wednesday, June 29th
Schedule
6:00 p.m. to 8:00 p.m
Workload
2 hours
Syllabus
- Validation measures.
- Hyperparameters tuning.
- Stacking and blending.
- Model selection.
- Overfitting.
Professor
- Zhiyu Quan.
Date
Thursday, June 30th
Schedule
6:00 p.m. to 8:00 p.m
Workload
2 hours
Syllabus
- K-means.
- Hierarchical clustering.
Professor
- Zhiyu Quan.
Date
Friday, July 1th
Schedule
6:00 p.m. to 8:00 p.m
Workload
2 hours
Syllabus
- Principal component analysis.
Professor
- Zhiyu Quan.
Date
Saturday, July 2th
Schedule
9:00 a.m. to 11:00 a.m
Workload
2 hours
Syllabus
- Explainability.
- Techniques.
Professor
- Zhiyu Quan.
Date
Tuesday, July 5th
Schedule
6:00 p.m. to 8:00 p.m
Workload
2 hours
Syllabus
- Analytical bias.
- Model adjustment with unbias information.
- Data and model governance.
- Ethical model documentation.
Professor
- Zhiyu Quan.
Date
Wednesday, July 6th
Schedule
6:00 p.m. to 8:00 p.m
Workload
2 hours
Organizing committee
Ana María Maldonado Ardila:
Director of Graduate Faculty of Economy.
Fabio Andres Gómez De Los Rios:
Director Master in Quantitative Finance.