Computational Data Analytics with Python

Roan G. W. Salgueiro

I built a predictive model using the Apprentice Chef, Inc. dataset from the pre-course assignment. The main objective of the assignment was to develop a model that predicts the continuous response variable, REVENUE. This required me to go through the steps of feature engineering, model preparation, variable selection, and model development.

To successfully complete the task, I utilize the knowledge and skills acquired during the course, especially in data science and machine learning. The goal was to build a model that accurately predicts the revenue generated by Apprentice Chef, Inc. and provides insights that can help the company make data-driven decisions. The completion of this assignment was a crucial step towards becoming proficient in using data to make informed business decisions.

In the following code, my logic is to analyze the data, treat in the dataset all anomalies (such as missing or null values and fill with 0 and correct column names). Then I run the correlation for all the numeric values as X against REVENUE Y to check what are the top correlations. In the next step, I develop new features and run OLS Statsmodels to check if the P-value has a signific statistic difference and check with one that has the most valuable value. Selected the best performance feature, I use the flow of scikit-learn: Instantiate, Fit, Predict, and Score.