Inventory Prediction Dashboard Project Data Scientist
Current- Developed dashboard to forecast manufacturing needs of new albums for the United States market by predicting the cumulative sales in 6 months after albums’ releasing - Feature Engineering: created and added new features such as time/sales difference, Google Trends- Fitted the time series models: auto-arima and VAR model in 18 different genres and applied sample test for each genre’s sales prediction - Built and visualized random walk model by predicting sales average and its confidence interval - Helped to apply machine learning methods (Decision Tree/ Ransom Forest/Xgboost) to set up the function to predict sales for each album
Jan 2023 - Present
