Time Series Mastery: Hands-On Workshops

Free with Premium Yearly Subscription

Overview

Time series forecasting is more than just predicting future trends - it’s a critical skill for industries ranging from finance to healthcare, retail, and beyond. Join us for a one-day virtual event packed with expert-led workshops designed to equip you with the latest AI-driven and classical forecasting techniques.

What’s on the agenda?

Optimizing Forecast Stability and Accuracy

30 min - Talk

In many business applications, forecasting models are optimized solely for accuracy, often leading to a common pitfall: month-over-month volatility in predictions, commonly known as the whiplash effect. This instability can erode stakeholder trust and disrupt planning processes, particularly in supply chain and demand forecasting. In this talk, we introduce a novel approach leveraging genetic algorithms to optimize both forecast stability and accuracy, creating a dynamically weighted ensemble that balances these competing objectives and delivering better accuracy than any single base model. By incorporating past model performance into our evolutionary framework, we iteratively evolve an ensemble that minimizes large forecast swings while maintaining or improving overall accuracy. We demonstrate how this method systematically adjusts model weights based on historical deviations and performance metrics, solving a key business challenge.

State of Foundation Models For Time Series Forecasting

30 min Talk

In this talk, we cover the state of foundation models in the field of time series forecasting. This approach is still relatively new, as we saw increasingly more models appearing since the start of 2024. First, we explore the core concepts of foundation models, such as pretraining, transfer learning and fine-tuning. Second, we take a look at the advantages and disadvantages of foundation models in time series forecasting. While they can speed up the modeling and inference process, they might also not be the best solution for a particular project, meaning that we must still have a certain expertise to use them correctly and compare them with other methods. Then, we explore some of the major contributions to the field, including TimeGPT, Chronos, Moirai and TimesFM. We quickly discover their architectures, their capabilities and their limitations. Finally, we see TimeGPT in action to demonstrate how a foundation model can be used and how it compares to traditional methods.

Forecasting the Future Using Time Series

1hr Workshop

Time series forecasting remains a specialty topic specializing in "predicting the future". Because of this, you really want to use a package that is tuned for your use case, and specialized to deal with the difficulties inherent in time series forecasting. I will share a simplified problem notation that helps you to survey available solution offerings, and succeed with time series packages in R and Python.

Leaning 1: time series software and terminology.

Learning 2: the importance of attributing and inferring hidden state in time series problems.

Learning 3: deploying reliable and effective forecasts in R and Python using standard open source packages.

Unlocking the Future with AI-Driven Time Series Forecasting

2hr Training

Time series forecasting is the science of predicting future events based on historical data, a practice with applications that permeate our daily lives. Consider demand and inventory planning, where forecasting enables businesses to anticipate customer needs, ensuring optimal product availability while minimizing costs.

Similarly, marketing strategies rely on forecasting to predict market trends, customer behavior, and campaign performance for effective resource allocation. In capital budgeting, forecasts provide insights into the future cash flows of investments, guiding informed decision-making.

Pricing optimization and predictive maintenance also depend on robust forecasting, whether it’s anticipating demand elasticity or preventing equipment failures through data-driven insights.

Training Outline:

The goal of this workshop is to demystify AI-driven forecasting, empowering participants to apply these tools for actionable insights in their unique domains. Through an interactive mix of lectures, live demonstrations, and practical exercises, attendees will gain both theoretical knowledge and hands-on experience with cutting-edge forecasting techniques. Whether you’re looking to refine your technical skills or explore how AI can unlock transformative insights, this workshop offers a practical gateway to mastering the future of forecasting.

New Frontiers in Time Series Foundation Models

30 min Talk

This talk explores the revolution of foundation models in time series, highlighting breakthroughs like TEMPO and TimeDiT. We'll examine their impact on forecasting, imputation, and anomaly detection across domains like energy, healthcare, and finance, and discuss future research directions.