Forecasting Models from A to Z

There is often confusion between forecasting methodologies and predictive modeling using supervised machine learning algorithms. While the latter relies on external information for its predictions, forecasting uses its own data.

This workshop aims to provide a comprehensive understanding of all forecasting methods and how to apply them for near-future predictions. It will cover basic models and then explore the evolution of various methods, enabling participants to use them effectively. Understanding all quality indicators will help participants select the best forecasting model for their businesses.

Workshop Overview

Learning Outcomes

• Compare forecasting with supervised machine learning.

• Learn how to select between forecasting models.

• Evaluate the relationship between the future and the past.

• Measure the impact of the past on the near future.

• Analyze all forecasting methods and their evolution.

• Develop all analytical models for estimation.

• Master the precision measures of models’ quality.

• Select the best forecasting model.

• Apply models with specialized software.

Detailed Course Schedule

• Day 1:

  • Linear and Polynomial trends

  • Exponential, Power, and Logarithm trends

• Day 2:

  • Averaging and Moving Averages

• Day 3:

  • Simple, Double, and Triple models in exponential smoothing

• Day 4:

  • Time Series

    ARIMA and Box Jenkins method

• Comprehensive colored PPT documents.

• Supervised ML vs. Forecasting approach.

• Stationary, Additive, and Multiplicative models.

• Proprietary tools solutions.

• Quality measures of forecasting models.

• "White Noise” data.

• Selecting the Fit model.

What will it be about?