Construction and Evaluation of Predictive Models Report (BUS5PA)

This report presents a comprehensive analysis of predictive modeling techniques using the R programming language. The assignment begins with data exploration and cleaning of the 'Cereals.csv' dataset, identifying variable types (continuous, ordinal, nominal), handling missing values, and visualizing data distributions using boxplots and histograms. Two primary datasets were used in this assignment. The first dataset, on breakfast cereals, was used to explore data cleaning and transformation techniques. The second dataset, on Toyota Corolla car prices, was used to construct and evaluate predictive models. The second part of the report focuses on building predictive models for car price prediction, involving feature selection, correlation analysis, and the construction of multiple regression models. The report then compares the performance of regression models with a decision tree model, evaluating their accuracy using metrics like RMSE. The report includes detailed tables, graphs, and statistical outputs to support the analysis and conclusions. The report compares the performance of various models, highlighting the best fit based on evaluation metrics and providing insights into the factors influencing car prices. The analysis demonstrates the application of predictive modeling techniques and the importance of model evaluation in real-world scenarios.