On Shelf Availability
Published:
About
This project was part of my curriculum at the Engineering School Télécom SudParis in collaboration with the startup Acuity-Data.
The goal was to analyse data from 200+ hyper-markets and apply machine learning algorithms on the cleaned data to predict On Shelf Availability. The algorithms involved standard Machine Learning methods such as regressions, random forest trees, but also Neural Networks.
Sadly, the data was too sparsed and poorly collected from the hyper-markers to reach satisfying results. Although it was a great case of handling real-world data related to retail.
Code Link
Due to the nature of the project, I have no right to share the code.
Stack
Numpy, Pandas, Matplotlib, Scikit-Learn, TensorFlow