1
Time-Series Analysis
Decompose sales data into trend and seasonality. Understand moving averages and autocorrelation using statsmodels.
Build a strong foundation in Data Analytics and AI using a case study of the global German brand Adidas. In this project, you will explore how raw market data transforms into actionable insights, identify macroeconomic patterns, and train predictive machine learning models to forecast retail revenue like a professional.
Just completed
Built a retail demand forecasting pipeline in Python comparing SARIMA and Linear Regression models using 9,600+ Adidas transactions. Formulated inventory recommendations to minimize Q4 overstocking risks. Graded and verified by ProoV.
Load and clean the Adidas US Sales dataset. Perform Exploratory Data Analysis (EDA) including time-series line plots of monthly revenue, regional bar charts, and product category breakdowns.
Resample daily data to monthly aggregates. Decompose time series into trend, seasonal, and residual components. Compute moving averages and plot autocorrelation (ACF) to identify lag patterns.
Train a Moving Average baseline, a Linear Regression with time features, and a Prophet model. Compare models using MAE, RMSE, and MAPE metrics. Visualize forecasts vs actuals.
Interact with the in-app Sales Dashboard to explore regional forecasts. Write an executive summary covering the strongest growth regions, product investments, and channel mix shifts.
Decompose sales data into trend and seasonality. Understand moving averages and autocorrelation using statsmodels.
Build robust forecasting models using Prophet and Linear Regression to predict regional sneaker demand.
Translate model outputs into actionable business decisions for inventory planning and channel investment.