Your task for this coursework is to construct classifier models to predict credit card loan defaults. You are given a csv file which contains 284808 observations and 29 features. Your goal is to use those features to predict the “Class” variable.
Question 1: Construct a logistic regression model for default prediction. Explain the workings of the model and evaluate your model performance on a suitable metric [15 marks]. What is the accuracy of your model and is this a good metric to use? [10 marks]
Question 2: Construct a random forest classifier model. Explain the workings and evaluate your model performance. [15 marks]. What are the hyperparameters for this model and how did you tune them? [10marks]
Question 3: Construct a support vector machine classifier model. Explain workings and evaluate model performance [15 marks]. What are the hyperparameters, how do you tune them and what happens if you change them? [10 marks]
Question 4: For each of the 3 models, can you improve the performance using techniques to deal with unbalanced data ? [15 marks]. Choose the best model in terms of performance (after hyperparameter tuning) and compare the performance of across all three models. Which model performs the best? [10 marks]
本网站支持 Alipay WeChatPay PayPal等支付方式
E-mail: email@example.com 微信号:vipnxx