CASE STUDY

REPLENISHMENT

  1. Sriya-AI can help match product ranges to variations in customer demand and optimize initial allocation, replenishment, and assortment planning
  2. The dataset used had 200,000 rows with 6 categorical and 9 numerical features
  3. Regression models were created using Lasso, Random Forest, XGBoost, RNN, and ANN algorithms with performance measurements obtained for feature sets of 4, 7, 10, and 12
  4. Average accuracy between 41% – 61% was achieved in predicting order demand, with major factors including category, checkout_price, cuisine, and base_price.