Would you like to know the future?

You can with time series - well almost!

In this hands-on workshop (limited seats to 30 people) we walk you through some of the basic approaches used to forecast times series data.

We begin with ARIMA models, and argue that techniques used their anticipated procedures used many machine learning algorithms. ARIMA models remain relevant, particularly when trends extend over months and years.

Next we take a quick look at Prophet, Facebook’s open source framework. Prophet provides a sophisticated framework for making and displaying forecasts quickly. It also provides several conveniences that make it easy to, for instance, include the effects of holidays in your analysis.

Finally, we look at RNNs, and particularly LSTMs, and how to implement them. These are very flexible tools, but finding the right parameters can be tricky. We can offer some suggestions.

We will provide the code to get started with each of these approaches.

18:00 food

18:30 Hands-on workshop

20:30 drinks

Looking forward to welcome you to our workshop,

Jurriaan and Jeff