Probabilistic Cash Flow Prediction based on historical transactions
This notebook will showcase a Cash Flow Prediction Engine, which aims to predict (in a probabilistic way) the cash inflows and outflows of the given categories each account of an SME based on historical transactions of SMEs.
The Engine’s objective and the nature of the task implied the necessary data transformation in time-series representation, enabling the experimentation with various general forecasting models or prevalent DL models. Thus, both resampling and aggregating the amount of the transactions based on each specific account and date have been applied to the data sample. For this task the DeepAR model was chosen as the main model for this forecasting task.
Creator: George Makridis
Github: https://github.com/GeorgeMakridis/Cas...
LinkedIn: / george-makridis-35608132
Dac Research Group:: https://dac.ds.unipi.gr
INFINITECH Marketplace: https://marketplace.infinitech-h2020....