In this chapter, we looked at machine learning and its different subcategories. We explored supervised, unsupervised, and reinforcement learning strategies, and looked at situations where each one would come in handy.
Looking into linear regression, we were able to find relationships between predictors and a continuous response variable. Through the train/test split, we were able to help avoid overfitting our machine learning models and get a more generalized prediction. We were able to use metrics, such as the root mean squared error, to evaluate our models as well.
By extending our notion of linear regression into logistic regression, we were able to then find association between the same predictors, but now to categorical responses.
By introducing dummy variables into the mix, we were able to add categorical features to our models and improve our performance even further.
In the next few chapters, we will be taking a much deeper dive into many more machine learning models and, along...