I choose to enroll in this course in an effort to gain more experience with applying machine learning techniques to other real world problems. As someone who already took, and loved, the primary machine learning course it made a lot of sense to apply those same skills to round them out further. Also given one of my bachelor’s degrees is in Economics and my personal interest in the stock market is made sense to further dive into machine learning for trading.
The first part of the course focused on utilizing Python Pandas, numpy, and scipy on stock data. There were a number of assignments to import data, clean it, manipulate it, and calculate various items on it.
The second part of the course was focused on building financial indicators, understanding market mechanisms, and doing technical analysis. Here I learned about CAPM, Bollinger Bands, and other indicators that can be used. In this part of the course we also analyzed various strategies that are used to generate a portfolio. For example, holding one stock and trying to micromanage it versus holding large numbers of stocks and letting it ride out the ups and downs.
The final section of the course was utilizing actual machine learning algorithms against portfolio data. Here I used KNN and linear regression algorithms in order to make predictions as to whether to buy or sell. Scikit-learn, another Python library, was leveraged in order to do some of these calculations.