Machine Learning in Finance

I wanted to link to an excellent tutorial on Precision vs Recall, and its harmonic mean F1 score  https://machinelearningmastery.com/precision-recall-and-f-measure-for-imbalanced-classification/

This post will describe an Information Theoretic approach to estimating a random variable. This approach again highlights how important Electrical & Electronics Engineering concepts are applicable to Capital Markets. 1. Caveats of correlation: Correlation is not a useful metric of co-dependency between two variables if: If there is a non-linear relationship between the variables If there are significant outliers that skew the correlation metric If the two variables do not jointly follow a bi-variate normal distribution If correlation is judged to be a useful metric instead of its usual form, it is worth using it in a modified form to judge the distance between two metrics 2. Shannon's Information-Theoretic approach: We need a metric that highlights the degree of randomness in a random variable. E.g. the degree of sunlight at a point throughout a 24 hour period is a random variable, but there is a low degree of randomness in this random variable. The entropy in a random variable...