Retail
Demand Forecasting & Inventory Simulation
MSc dissertation project: forecasting demand with five models, then simulating how each forecast would change stock levels and costs.
[Add your headline result, e.g. "Best model cut forecast error by X% vs ARIMA"]
01 · The problem
[What business problem were you solving? Who loses money or time when demand is forecast badly?]
02 · The data
[Dataset: source, size, time span, and the cleaning it needed.]
03 · The approach
Compared ARIMA, Random Forest, XGBoost, LSTM and N-BEATS forecasts, fed them into an inventory simulation, and built a Flask web app and a Power BI dashboard so non-technical users could explore the results.
04 · The results
[Which model won, by how much, and what it means for stock decisions.]
Use this space for the full case study: charts, screenshots of the Power BI dashboard, what surprised you, and what you'd do next.