This repository is the project space of the team ALMA2020 - Mishmash Online Hackathon.
The sales prediction tool is developed thoughtful planning of data, user-experience, performance, and stability. The tool designed is integrated with the Machine-Learning model for the purpose of prediction.
The machine learning model is developed with Microsoft Azure - Machine Learning Toolsuite, to ensure the development of best performance models and algorithms. A total of 45 models were compared for various performance metrics and the best among the models is interfaced with the tool.
The frontend of the tool is developed with Javascript and interface with Python in the backend. This is facilitated by the Flask Framework. The template for GUI was developed with Jinja 2 Template Language.
The implementation of the machine learning model with such high accuracy of 99.262% and online deployment of the prediction tool, provides state-of-the-art tool infrastructure to predict sales for an extended period of time.