LentaHack🌻 DeepHunch, 2020

Skoltech,HackLab Course

These notebooks are provided in frames of Lenta Hackathon during HackLab Course at Skoltech. The descriptions of the presented data can be found below.

1. Notebook Predprocess_lenta.ipynb identifies clients whose last visit was earlier than august,which may be a sign that they left the Lenta shop.The function 'find_7goes_busket(client)' helps to determine the goods that were stably bought by the client. This hepls us to identify the preferred goods of the clients who left Lenta.
2. The two notebooks Visualization.ipynb and partially Vizualization_and_Time_Series.ipynb consist of a short visualization of the datasets. As the dataset was huge we splited this task.

3. Also the second notebook Vizualization_and_Time_Series.ipynb provides simple illustrations of time series predictions with ARIMA model. The result showed that for effective prediction more complicated model is needed.
4. Notebook Time_series_clustering_model.ipynb contains time series clusterization for bills time series of clients. The result is that due to low computational resources, the testing remains unfinished.