Applied Modeling and Inferential Statistics

This unit develops applied supervised and unsupervised modeling skills in R and Rattle (decision trees, SVM, neural networks, clustering) and reinforces data visualization and reproducible workflow practices. It also introduces fundamental inferential statistics (t-tests, z-tests, variance assessment, medians and summarizing) to evaluate model results and real-world datasets through project-based activities.

Learning Objectives

• Apply supervised and unsupervised machine learning methods (decision trees, support vector machines, neural networks, and clustering) to real datasets using R and Rattle and interpret their performance metrics
• Analyze and visualize dataset features and model outputs using ggplot and R session tools, including package management and reproducible sidebar/workflow practices
• Demonstrate the ability to perform and report basic inferential statistical tests in R (one- and two-sample t-tests, z-tests, and tests for equal variances) and summarize central tendency and distributional characteristics
• Evaluate model choices and preprocessing decisions through project-based assessments (e.g., glass identification, mushroom/house/banknote datasets) and recommend improvements based on statistical evidence and validation results