---

Open source statistical toolkits have progressed tremendously over the last decade. In this book Nicolas Vandeput demonstrates that these toolkits are more than enough to address real-world forecasting challenges as found in supply chains.

Using data science in order to solve a problem requires a scientific mindset more than coding skills. Data Science for Supply Chain Forecasting contends that a true scientific method that includes experimentation, observation and constant questioning must be applied to supply chain as well. The first part of the book is focused on statistical "traditional" models and the second on machine learning. The various chapters are focused either on forecast models or on new concepts (overfit, underfit, kpi, outliers). The book is full of python examples to show the reader how to apply these models him/herself.

This is a book for practitioners focusing on data science and machine learning and demonstrates how both are closely interlinked in order to create an advanced forecast for supply chain. Through its hands-on approach, it is accessible to a large audience of supply chain practitioners.

---

I Statistical Forecast

Moving Average

Forecast Error

Exponential Smoothing

Underfitting

Double Exponential Smoothing

Model Optimization

Double Smoothing with Damped Trend

Overfitting

Triple Exponential Smoothing

Outliers

Triple Additive Exponential smoothing

II Machine Learning

Machine Learning

Tree

Parameter Optimization

Forest

Feature Importance

Extremely Randomized Trees

Feature Optimization

Adaptive Boosting

Exogenous Information & Leading Indicators

Extreme Gradient Boosting

Categories

Clustering

Glossary

---

"I had a chance to review the manuscript. It is a very good book. For the supply chain managers out there, you should read at least the first few chapters, and then have others on your team read the rest of it and act on it ... you can have close to state-of-the-art forecasts with a minimum of effort.... This book closes the coffin on vendors who are selling only a handful of forecasting models."

-- Joannes Vermorel , Founder and CEO, Lokad

---

Nicolas is a Supply Chain Data Scientist specialized in Demand Forecasting & Inventory Optimization. He always enjoys discussing new quantitative models and how to apply them to business reality. Passionate about education, Nicolas is both an avid learner and enjoys teaching at universities including the University of Brussels; he teaches forecast and inventory optimization to master students since 2014. He founded SupChains in 2016 and co-founded SKU Science-a smart online platform for supply chain management-in 2018.

---