| Content | This meetup will conclude the first stage (which mainly focuses on supervisor learning) of our meetup series on Machine Learning in Quantitative Finance. We will first have a group discussion on the selected case study (given as follows) and then a drink afterwards.In the group discussion, we will discuss how to apply supervised learning techniques covered in the previous sessions to solve real world problems, and how to handle some commonly raised practical issues in data challenges, e.g. missing data, outliers, etc.Through the discussion of hands-on ML projects, we will share the experience of using standard python packages for EDA, deep learning etc. In addition, this will be a great opportunity for one to participant in Kaggle Competitions, that shares many commonalities with many other types of data challenges.You are warmly invited to this event and to celebrate all we have completed in the first chapter of the meetup series, even though you are too busy to complete the exercises beforehand.This may be the last event in the recent time and we sincerely hope to see all of you.Topics of Case StudyData Challenge 1: Home Credit Default RiskCan you predict how capable each applicant is of repaying a loan?This task is a recent popular Kaggle Competition, in which more than 7000 teams have participated. The goal is to use historical loan application data to predict whether or not an applicant will be able to repay a loan. We can model the task using a standard supervised classification model.The description of the task can be found via the following link (https://www.kaggle.com/c/home-credit-default-risk.). To start with, you are suggested to take a look at the IPython notebook by Will Koehresen to gain an initial understanding of the data. https://www.kaggle.com/willkoehrsen/start-here-a-gentle-introductionData Challenge 2: Volatility ForecastThe task of this data challenge is to use the securities’ past volatility and price information to predict their future volatility and to control portfolio risk. Details can be found in the following link https://challengedata.ens.fr/en/challenge/34/volatility_prediction_in_financial_markets.htmlIf you have difficulty in opening the webpage and download the data, please find it from the link: https://www.dropbox.com/sh/7403m207xjhr2fa/AAACHE73J2HyqJogInuNJQA-a?dl=0. Here we provide the input and output data in the training set, and the input data in the testing data set, but keep the output of the testing dataset for evaluation. The evaluation metric of the output is defined as the mean absolute percentage error. You are expected to submit your estimated output for the testing data set to us.Link: https://www.eventbrite.co.uk/e/meetup-for-machine-learning-in-quantitative-finance-tickets-48603145316Passowrd: MachineLearningQuantFinance1 |