The goal ist to develop a comprehensive crm toolset, which focuses on revenue

Every data scientist dreams to predict future events - me too. But the world of time series forecast is much darker and bitter than expectet. Therefore this project will predict various values for a basic company. And, as the world is rarely 1-dimensional, the focus will be on predicting multiple attributes.

1. Univariate time series forecasting
2. Multivariate time series forecasting
3. Adding More Attributes and Keras LSTM

1. Univariate time series forecasting Forecasting revenue for the next four periods and make the prediction availble via Rest API

2. Multivariate time series forecasting Base the revenue prediction on multiple attributes and detect correlations, for example the importance of growth markets

3. Adding More Attributes and Keras LSTM Calculate customer KPIs, like RFM and CLV, and use a neuronal network to make revenue predictions

Credit for the great northwind database goes to conwid/awesome-northwind