A grocery store chain is planning a significant expansion. The objective of this project is to use multiple analytical techniques to provide recommendations on where and how to expand and deliver solutions to complex business problems.
This Project has 3 main parts:
The first part is to analyse the existing stores and the figure out how many formats or segments there are. This part is mainly done by unsupervised learning and I used K-means as the Clustering Algorithm.
The second part is to predict which of the new 10 stores fall in each segment. This part is about using supervised learning algorithms and I chose to use Decision Tree, Random Forest and a boosted model (AdaBoost base in Python and Boosted Model in Alteryx).
The third and last part is about forecast for the next 12 months which will be the Total Sales of Produce for the new and existing stores. In this part I used Exponential Smoothing (ETS) and Autoregressive Integrated Moving Average(ARIMA) Models to forecast the produce sales.
This project was made firstly in Alteryx Design and then in Python, some changes were required, but the goal was achieved in both cases.
The project report for the Alteryx solution in the "report_alteryx.docx" file and for the Python Solution in the "report_python.docx" file.