Training: Forecasting basics with real use cases.
Two-day, hands-on workshop: Learn forecasting fundamentals and apply them directly to real planning problems – from data preparation and method selection to evaluation and decision integration. Ideal for business and data/analytics teams.
About the Forecasting Workshop
Learn the fundamentals of forecasting in two days and how to apply them directly to your planning processes. The content is based on 10+ years of experience building production forecasting solutions across industries – with proven approaches so you can skip common pitfalls.
On day two we work on participants' questions ("Bring your own use case") and apply what we learned to real data. Programming skills are useful but not strictly necessary.
The Key Details
2 days
Recommended with 2–3 weeks between sessions
Interactive workshop
Theory plus hands‑on
€2,700 per person
plus VAT; volume discounts on request
What you'll take away
Data understanding & preprocessing: You will learn how to prepare raw data, define a suitable dataset, and identify patterns such as seasonality, trends, and outliers.
Methods & drivers: You will understand the differences between statistical methods, machine-learning approaches, and foundation models – and when to use which.
Evaluation & transfer: You will learn how to assess forecasts and address common data issues (e.g., missing values, assortment changes).
Business integration & impact: You will learn how to turn forecasts into decisions – for example through human-in-the-loop processes and traffic-light systems.
Agenda: four hands-on modules
From fundamentals to applying them to your own use cases.
Module 1: Data definition & preparation
- Define suitable data structures and clean raw data
- Identify patterns: trends, seasonality, level shifts, time‑series types
Module 2: Forecasting methods & evaluation
- Model landscape: naive baselines, models with drivers (promotions, weather), ML approaches
- Evaluation & backtesting strategies aligned with your business KPI
Module 3: Practice: Bring your own use case
- Live analysis of selected use cases – discuss and translate questions into concrete solutions
- Handle typical hurdles: data gaps, structural breaks (e.g., assortment changes), noisy data
Module 4: Process integration & outlook
- Business impact: operational constraints (e.g., order quantities) and decision aids (e.g., traffic‑light checks)
- What's next: how GenAI/LLMs can support analysis and external event detection
Prerequisites
- Domain knowledge: Understanding of your data and processes.
- Tech: Basic technical understanding is helpful; deep programming skills are not required.
- Bring your own data: We welcome your use cases. For live code analysis, please send a data sample and a concrete question at least one week in advance.
- Equipment: Your own laptop; either local Python + Jupyter, or browser + Google account (for Colab).