Logistic Regression with PyTorch

We learned about linear regression in the last post, and now we move to logistic regression. As we cannot use linear regression for a classification task, we use logistic, which is an extension to linear regression for a classification task. Logistic regression is a good approach to problems that require probability as output. Suppose we create a model to predict a striker's probability when he is playing an away match. If a model predicts a p(score Away_Match) of 0.05, then overall away matches, the striker will score approximately 1 goal.