Time Series Analysis

Course Overview

Type of course: classroom training, duration: 2 days

The two-day course Time Series Analysis teaches you what needs to be considered in the analysis and utilization of time series based data and how the data must be prepared in order to be processed with machine learning methods.

Almost all data sources have a time reference which needs to be taken into account in the analysis. This time reference can provide valuable information but can also give a false indication of correlations and must therefore always be kept in mind when creating predictive models. We start by presenting validation methods which take the time reference into account and which provide a realistic estimate of the expected quality of forecasts in situations of this kind, even where problems without a time reference are concerned.

Once participants have been familiarized with the underlying problems, the field of time series analysis is first defined and divided into the three areas:

  • Classification of entire time series,
  • segmentation of time series and
  • time series forecasting, i.e. predicting the next expected values.

Following this, possible problems and their solutions are discussed. For this purpose, various pre-processing methods are presented allowing the problem to ultimately be solved with the known machine learning methods. The course is structured to provide constant alternation between the study of theoretical basics and proven best practices and the practical application of knowledge acquired. Participants will form a data science team that completes the tasks set by the course instructor together. The exercises are carried out on the participants’ personal laptops, meaning they can refer back to the content as a starting point for solving their own challenges after the training course.

Course Objectives

The skills acquired by participants of the training course include:

  • Recognizing time reference of problems
  • Validation of predictive models in a context with time reference
  • Differentiating between time series classification, segmentation and forecasting
  • Feature extraction on time series
  • Segmentation of time series
  • Forecasting of time series

Previous Knowledge Required

For this training course, you will need the knowledge from the previous courses Basics 1 & 2. If you have already acquired equivalent knowledge from comparable training courses, please contact us.

Target Group

Data scientists, advanced analysts


Participants will receive a certificate of attendance.


€ 2000 per participant (plus VAt)

Information about arrival, accomodations and meals.