Business Analytics Projects
These are all the business analytics projects/coding assignments I engaged in through my first semester in the MS in Business Analytics program. These projects highlight my competencies in carrying out analyses and building prediction models using Decision Trees, k-Nearest Neighbors, and Logistic Regression. Furthermore, it also includes analyses I ran to evaluate the said models and pick the models that perform the best. Additionally, I also engaged in transformation and feature engineering to prepare my data for all the assignments.
The final project - Red Wine Quality Prediction Model - brings together all of these techniques I have learned through the semester to put it together and build a recommendation model for red wines based on different features such as its acidity, pH level, sulfur dioxide levels, etc. In this project, I engaged in feature engineering to create a new variable called 'Molecular Sulphur Dioxide' that recognizes the relationship between sulphur dioxide amounts in a wine and the pH level of the wine. This interaction variable increases complexity in the model but it also improves the model vastly when using models that are tolerant of complexity. Lastly, I also balanced the target variable 'Wine Quality' as it was imbalanced in terms of class distributions and used ROC and AUC to evaluate the model for its efficacy.