APPLICATION OF MACHINE LEARNING TO DERIVATIVE PRICING

PROJECT DETAILS

This project will be focusing on derivative pricing using classical financial mathematics methods as well as on novel approaches of applying machine learning (ML) techniques to valuation. Students will start from understanding market data, risk factors and derivative valuation. A simple Monte Carlo (MC) simulation engine will be built by the students at this stage. Then two popular ML techniques, Gaussian Processes and Deep Learning, will be applied to derivative pricing. Finally, a thorough analysis and comparison between ML techniques and classical methods will be conducted. The MC simulation engine built in the beginning of the project will be used to generate training data. We will focus on pricing basket options and, time permitting, on some other type of derivatives. Results of different ML approaches will be benchmarked against the market, classical financial mathematics approaches, and similar published studies. Python will be used as a main programming language. After completing the project the students will have a good understanding of financial data, MC simulations, derivatives, risk sensitivities, valuation, as well as ML and its application to derivatives pricing.

POTENTIAL EMPLOYMENT OPPORTUNITIES

Front Office Quants
Risk Management Quants
Data Management Office Quant
Data Vendor Quant
Data Scientist in a Financial Institution

PROJECT DURATION & COST

Program Duration: 10 months
Upfront Cost: $0
Upon Securing Employment: $9,999

MENTORS

Dr. Dmitry VyushinDIRECTOR OF MARKET RISK MODELS, RBC

Dr. Dmitry Vyushin is a Director in Non-Trading Book Risk Modelling Team of Market Risk Department at the Bank of Montreal. He is currently developing a new Enterprise VaR system that covers Trading, Treasury, Insurance, and Pension portfolios. He also holds a PhD in Physics from the University of Toronto.

Dr. Markiyan SlobodaVICE PRESIDENT CAPITAL MARKET, BMO

Dr. Markiyan Sloboda finished his PhD in Computer Science from the University of Guelph. Currently, Dr. Sloboda is working as a Vice President in BMO Capital Markets, responsible for developing equity vanilla pricing models and supporting the volatility trading desk in their everyday activities. During his tenure at TD and BMO, Dr. Sloboda was also teaching for 3 years at the Ted Rodgers Business School, Ryerson University.

CUSTOMER PROFITABILITY USING RETAIL DATA

PROJECT DETAILS

In this project you will explore retail data to build Machine Learning models and predict customer profitability as well as Lifetime Value (LTV). You need to cover all the steps of a generic data science project: data exploration, visualization, model selection, training, validation, fine-tuning, testing, etc. Then, after building your benchmark model, you will explore a wide range of Machine Learning algorithms for a complete comparison between results.

At the end, you will deploy your final model to Google Cloud Platform and make it accessible to other users. Through this process, you will gain invaluable skill sets of a full stack Data Scientist.

POTENTIAL EMPLOYMENT OPPORTUNITIES

Data Scientist in Retail Companies
Data Scientist in Consulting Firms
Data Scientist in E-Commerce Companies

PROJECT DURATION & COST

Program Duration: 5-7 months
Upfront Cost: $0
Upon Securing Employment: $8,999

MENTOR

Dr. Mahdi ShahbabSenior Data Scientist, RBC

Dr. Mahdi Shahbaba is a Senior Data Scientist in Marketing Science team at RBC. He is currently working on Machine Learning models for product recommendation, pricing, and customer profitability. Prior to RBC, he was a Lead Data Scientist in Digital Factory and Decision Sciences departments at Scotiabank, where he developed ML models for fraud prevention and marketing campaigns.

He has a PhD in Electrical and Computer Engineering from Ryerson University with the main focus on Unsupervised Machine Learning.

MARKETING CAMPAIGNS

PROJECT DETAILS

In this project you will build Machine Learning models to predict client response to marketing campaigns. The trade-off between profit and response, marketing channel selection and cost minimization are some of the challenges that you have to overcome in any marketing campaign. You need to cover all the steps of a generic data science project: data exploration, visualization, model selection, training, validation, fine-tuning, testing, etc. After building your benchmark model, you will explore a wide range of Machine Learning algorithms for a complete comparison between results. At the end, you will deploy your final model to Google Cloud Platform and make it accessible to other users.

POTENTIAL EMPLOYMENT OPPORTUNITIES

Data Scientist in Financial Institutions
Data Scientist in Insurance Companies
Data Scientist in Telecom Companies

PROJECT DURATION & COST

Program Duration: 5-7 months
Upfront Cost: $0
Upon Securing Employment: $8,999

MENTOR

Dr. Mahdi ShahbabSenior Data Scientist, RBC

He has a PhD in Electrical and Computer Engineering from Ryerson University with the main focus on Unsupervised Machine Learning.

Current Projects

CURRENT PROJECTS

APPLICATION OF MACHINE LEARNING TO DERIVATIVE PRICING

CUSTOMER PROFITABILITY USING RETAIL DATA

MARKETING CAMPAIGNS