Join our workshop at QuantMinds International with John Hull, Maple Financial Professor of Derivatives & Risk Management, Joseph L. Rotman School of Management at University Of Toronto
It is now generally recognized that machine learning will have a huge impact on risk management and quantitative analysis in the years to come. This workshop will not convert you into a data scientist! But it will put you in the position where you can work productively with data scientists. You will understand the tools they use and know the right questions to ask. Some of you I hope will be inspired to learn more and become data science experts. (A number of the students at the Rotman School of Management, University of Toronto, who took courses with me that are similar to the workshop, have done this.)
The first part will introduce the different types of machine learning (supervised, unsupervised. reinforcement). It will stress the importance of testing models out of sample and explain why a data set should be divided into a training set, validation set and test set when used for prediction. It will also cover a number of different tools that have proved to be effective for clustering. These tools have enabled companies to understand their customers better and communicate with them more efficiently.
The second part will move on to consider how machine learning is used for prediction. What is the price of a house with certain features in a particular London suburb? What is the probability that a particular borrower will default? We will discuss how data scientists have extended the range of applicability of well known tools such as linear and logistic regression. We will also cover important machine learning tools such as support vector machines, decision trees, random forests and neural networks.
The third part (after lunch) will be devoted to reinforcement learning. This is an important “trial and error” technique in machine learning. It is designed for situations where a series of decisions have to be made in an environment that is changing unpredictably.
It has seen many applications in finance. It can be used to develop hedging strategies for derivatives, manage portfolios of assets, and determine the best way of buying or selling a large block of shares. It is a very powerful tool.
AlphaZero, a computer program which was given nothing more than the rules of chess, used RL to become better than any human chess player in a matter of hours!
The fourth part of the workshop will cover natural language processing. This is a really exciting area of machine learning which will have a big influence on all our lives in the future. Great progress is being made in designing algorithms to understand both written and spoken language. It is making professional translators obsolete! The fourth part of the workshop will also cover ML explainability, which becoming an increasingly important topic.
Maple Financial Professor of Derivatives & Risk Management
