Companies have huge amounts of data but they do not really know what or how they should use this data. Often times it is very expensive to make accurate predictions. We provide a very accurate prediction to how a company should expand, predicting what products each store should focus on, give personalised recommendations for each customer, by building user friendly and intuitive dashboards. Poorly placed stores and irrelevant product offerings lead to reduced customer visits. Our solution optimizes inventory and store placement to drive higher footfall and revenue.
Finding datasets was very difficult, but we solved this by generating dataset using Faker, and after that augmenting the data to prevent overfitting and reduce variance in data.
During training we faced low accuracy and sometimes overfitting. We solved this by using XGBoost instead of Decision Trees.
Tracks Applied (1)