PLA-43136 Time Series Analysis, 5 cr

Additional information

The course can be completed with two different implementation methods. The course provides traditional lectures and conducts assignments. However, the teaching material is available on Moodle platform, so it is also possible to complete the study period independently of time and place throughout the academic year. If the student intends to complete the course outside of the lecture period, he / she should contact the person in charge, jari.j.turunen (at) tut.fi, for obtaining course IDs.

Person responsible

Jari Turunen

Learning Outcomes

After completing the course, students are able to search for phenomena from different time series (sensor data, historical time series, economic and productive time series), clean data and search dependencies between time series related to the same phenomenon, such as temperature and atmospheric pressure. The course is completed by assignments.

Content

Content Core content Complementary knowledge Specialist knowledge
1. Reduction of measurement noise in time series (eg average / weighted average), different filtering options   Visualization of the phenomena both in time and frequency domains  Understanding the Dependence Between Time and Frequency Domains 
2. Understanding the concept of internal dependencies (correlation) of the time series, understanding the dependencies between several types of measurements  Implementation of correlation search between the measurement sets.   
3. Change rate (differential) and automatic phenomena extraction from data and isolation of significant data from data  Understand how phenomena can be searched automatically   
4. Prediction of data by linear correlation, estimation of missing data  Knowing the AR model well, and knowing where and when to use it  Knowing another prediction models such as AR,MA and AR(I)MA linear models well and have ability to use them 

Instructions for students on how to achieve the learning outcomes

Approved assignments

Assessment scale:

Numerical evaluation scale (0-5)

Partial passing:

Completion parts must belong to the same implementation



Correspondence of content

Course Corresponds course  Description 
PLA-43136 Time Series Analysis, 5 cr PLA-43131 Time Series Analysis, 5 cr  

Updated by: Baggström Minna, 16.10.2018