Time series analysis

News

• The results of the take-home exam are now posted on the notice board. The project reports will be graded within two weeks.
• There will be additional office hours on Thursday 10/1 and Tuesday 15/1 at 11-12.
• The take home exam has now been published (see below). Best of luck!
• There will be additional office hours on Friday 14/12 at 11-12 and 14-15.
• Please note that all predictions should be presented in the original data domain (in this case temperature), not in any differentiated domain. Also, do not forget to remove samples corrupted due to the lack of initial states when filtering data.
• There was a transpose too much in lbpTest. The matlab code for the book has now been updated.
• There will be no office hours on Wednesday 12/12, but instead on Thursday 13/12 at 1 pm.

Course contents

Time series analysis concerns the mathematical modeling of time varying phenomena, e.g., ocean waves, water levels in lakes and rivers, demand for electrical power, radar signals, muscular reactions, ECG-signals, or option prices at the stock market. The structure of the model is chosen both with regard to the physical knowledge of the process, as well as using observed data. Central problems are the properties of different models and their prediction ability, estimation of the model parameters, and the model's ability to accurately describe the data. Consideration must be given to both the need for fast calculations and to the presence of measurement errors. The course gives a comprehensive presentation of stochastic models and methods in time series analysis. Time series problems appear in many subjects and knowledge from the course is used in, e.g., automatic control, signal processing, and econometrics.

Higher education credits: 7,5 Level: A
Language of instruction: This course may be offered in English (if non-Swedish speaking students are attending).
Recommended prerequisites: Basic courses in probability and statistics, as well as stationary stochastic processes.
Literature: Andreas Jakobsson, Time Series Analysis and Signal Modeling. The lecture notes can be bought from the course secretary.
Time: A detailed schedule for the course can be found here. Tutorials will be held on Tuesdays and Fridays at 10.15-12.
Office hours: The lecturer will have office hours in MH:225 on Mondays 13-14 and Wednesdays 11-12 (until 19/12, except 12/12). Without appointment, please respect these hours.
Examination: The course examination consist of mandatory computer exercises, a take-home exam, as well as a project. As a part of the examination, a detailed project report should be handed in, as well as the result disseminated in an oral presentation. Project examination will take place on 17/12, at 10-12, or 18/1, at 15-17, in MH:309A (choose either of the times). The presentation material should be handed in no later than at the start of the presentation. The take home exam is due on 22/12, at 11.00.

Course material and schedule

• Course program
• Problem solutions
• Matlab files for the computer exercises: zip, as well as for the textbook
• Lecture notes (published as we go)
  • Lecture 29/10 - A:B - Introduction and overview. Multivariate random variables.
  • Lecture 31/10 - M:E - Multivariate random variables. Stochastic processes.
  • Tutorial: 30/10 - MH:228 - Problems: 2.1, 2.2, 3.1-3.4
  • Tutorial: 2/11 - MH:227
  
  
  • Lecture 5/12 - A:B - Stochastic processes.
  • Lecture 7/11 - M:E - Stochastic processes. Identification.
  • Tutorial: 6/11 - MH:333 - Problems: 3.5-3.10, 3.12-3.15
  • Tutorial: 9/11 - MH:228
  
  
  • Lecture 12/11 - A:B - Identification.
  • Lecture 14/11 - M:E - Estimation.
  • Tutorial: 13/11 - MH:228 - Problems: 4.1-4.4
  • Tutorial: 16/11 - MH:228
  
  
  • Lecture 19/11 - A:B - Model order selection.
  • Lecture 21/11 - M:E - Residual analysis.
  • Tutorial: 20/11 - MH:228 - Problems: 5.1-5.5, 5.8, 5.10-5.12
  • Tutorial: 23/11 - MH:228
  
  
  • Lecture 26/11 - A:B - Prediction.
  • Lecture 28/11 - M:E - Multivariate time series.
  • Tutorial: 27/11 - MH:228 - Problems: 6.1-6.8
  • Tutorial: 30/11 - MH:228
  
  
  • Lecture 3/12 - A:B - Recursive estimation. State-space models. The Kalman filter.
  • Lecture 5/12 - M:E - The Kalman filter. Project discussion.
  • Tutorial: 4/12 - MH:228 - Problems: 7.1-7.2, 8.1-8.2
  • Tutorial: 7/11 - MH:228
  
  
  • Lecture 10/12 - A:B - Reserve.
  • Tutorial: 11/12 - MH:228 - Problems: 8.3-8.8
  • Tutorial: 14/12 - MH:332B
  
  
• Project
  
  • Project description
  • Data file
  Project examination will take place on 17/12, at 10-12, or 18/1, at 15-17, in MH:309A (choose either of the two times). The presentation material should be handed in no later than at the start of the presentation.
  
  
• Take-home exam
  The take-home exam is available here. The exam is due on 22/12, at 11.00, but can only be submitted via email after 21/12, at 12.00.
  

Computer exercises

Please sign up for the computer exercises (the list will be posted on the notice board).

Exercise 0	Do this one on your own!
Exercise 1, prediction	27/11, 15-19, MH:230	28/11, 8-12, MH:230,	29/11, 8-12, MH:230/231
Exercise 2, estimation and validation	4/12, 15-19, MH:231	5/12, 8-12, MH:230	6/12, 8-12, MH:230/231
Exercise 3, recursive methods	11/12, 15-19, MH:230,	12/12, 8-12, MH:231,	13/12, 8-12, MH:230/231

Suitable further courses

• FMS 110/MAS 222 Stationary and non-stationary spectral estimation.
• FMS 110/MAS 222 Non-linear time series analysis.
• FMS 161/MAS 229 Financial statistics.
• FMS 170/MAS 232 Valuation of derivative assets.
• Also, have a look at the available thesis projects.