Exam PA Details

The primary statistical software used for our analyses is R. Only a surface-level ability to recognize certain code would be needed.

Acquainting ourselves with the data is the next vital step. We learn how to understand and summarize the data with a few numerical and graphical techniques. More often than not, we then discover that the data is not ready to be analyzed with models, so steps are taken to clean the data. We begin by explaining our game plan for analyzing data throughout this course.

With the data adequately prepared, we can commence running predictive models. Once we understand the goal of statistical learning and its essential terminology, we study and ponder a variety of parametric models. This encompasses multiple linear regression, stepwise selection, regularization, and generalized linear models.

With parametric models behind us, we now study a variety of non-parametric models. These models have less assumptions about the model and data. Non-parametric models comprise of decision trees, random forests, and boosting.

Time to shift gears to learn about unsupervised learning. We employ unsupervised learning to create new features to be used in predictive models. We explore principal components analysis (PCA), k-means clustering, and hierarchical clustering.

The efficacy of your analysis hinges on the quality of your communication. Besides addressing the items to document in your report for submission, we discuss ways to be cognizant of and strategize around time and special instructions.