Machine Learning with Minitab Predictive Analytics

This comprehensive course is designed to provide a detailed understanding of the basics of machine learning using Minitab, with a focus on supervised learning. The course covers the fundamental concepts of regression analysis and binary logistic classification, including how to evaluate models and interpret results. The course also covers tree-based models for binary and multinomial classification.

The course begins with an introduction to machine learning, where students will gain an understanding of what machine learning is, the different types of machine learning, and the difference between supervised and unsupervised learning. This is followed by an overview of the basics of supervised learning, including how to learn, the different types of regression, and the conditions that must be met to use regression models in machine learning versus classical statistics.

The course then delves into regression analysis in detail, covering the different types of regression models and how to use Minitab to evaluate them. This includes a thorough explanation of statistically significant predictors, multicollinearity, and how to handle regression models that include categorical predictors, including additive and interaction effects. Students will also learn how to make predictions for new observations using confidence intervals and prediction intervals.

Next, the course moves onto model building, where students will learn how to handle regression equations with "wrong" predictors and use stepwise regression to find optimal models in Minitab. This includes an overview of how to evaluate models and interpret results.

The course then shifts to binary logistic regression, which is used for binary classification. Students will learn how to evaluate binary classification models, including good fit metrics such as the ROC curve and AUC. They will also use Minitab to analyze a heart failure dataset using binary logistic regression.

The course then covers classification trees, including an overview of node splitting methods such as splitting by misclassification rate, Gini impurity, and entropy. Students will learn how to predict class for a node and evaluate the goodness of the model using misclassification costs, ROC curve, Gain chart, and Lift chart for both binary and multinomial classification.

Finally, the course covers the concept and use of predefined prior probabilities and input misclassification costs, and how to build a tree using Minitab. Throughout the course, students will gain hands-on experience applying the concepts learned in real-world scenarios.

Overall, this course provides a thorough understanding of machine learning basics using Minitab, with a focus on supervised learning, regression analysis, and classification. Upon completion of this course, students will have the knowledge and skills to apply supervised machine learning techniques to real-world data problems.

Who this course is for:

• This course is designed for students with a basic statistics background who are new to machine learning and want to gain practical skills in this field. No programming experience is necessary, but the course will introduce you to the advanced use of Minitab's menu-driven interface. Machine Learning is a multi-disciplinary field, often only to be learned in more depth over several books and courses, but this course is the perfect first learning resource.