Forecast Visualisations
Key outputs from the analysis are displayed below.
Forecast values were compared against actual observations to evaluate model performance and forecasting accuracy.
Forecasting demand using historical sales data for The Alchemist and The Very Hungry Caterpillar
This project explores a range of forecasting techniques, from classical statistical approaches through to machine learning models, using historical book sales data.
By analysing trends, seasonality and demand patterns, the aim was to determine whether future sales could be forecast accurately enough to support business decision-making.
By analysing sales data from Nielsen's dataset, this project investigated whether identifiable trends and seasonal patterns could be leveraged to generate accurate forecasts.
Reliable forecasts can help organisations improve procurement planning, inventory management and stock replenishment decisions.
The analysis focused on two books:
Historical sales data was cleaned, transformed and prepared for time series analysis using Python.
The forecasting workflow followed a structured analytical process:
Key outputs from the analysis are displayed below.
Forecast values were compared against actual observations to evaluate model performance and forecasting accuracy.
Forecasting enables organisations to make more informed inventory and procurement decisions.
Improved demand planning can reduce stock shortages, minimise excess inventory and support more efficient resource allocation.
This project demonstrates how statistical forecasting techniques can transform historical sales data into actionable business intelligence.
Full notebooks, preprocessing steps, model development, evaluation and supporting documentation are available within the project repository.
View Repository