Topic outline

  • General

    This course is an introduction to time series analysis. Course topics include linear regression model and its diagnostics, central concepts of weakly stationary processes, ARIMA-models and their properties, stationarity of ARIMA-models, forecasting with ARIMA-models, Kalman filter, and introduction to dynamic regression models. 

    Software R is used in the exercises of the course. If you have never used R before, you could consider attending MS-EV0002 Introduction to R programming: 

    https://mycourses.aalto.fi/course/view.php?id=37985

    After passing the course the students can analyse and forecast time series using regression models and ARIMA-models. Students are able to apply linear regression model to analyse and forecast dependent variable under the model assumptions. In addition, the students are able to conduct diagnostic tests to validate the model assumptions. Students are familiar with the concept weakly stationary processes and they understand the most important related concepts including the autocorrelation function, partial autocorrelation function, and the spectral function. Students are also able to apply these functions in analysing real time series, for example, in recognizing seasonal fluctuations. After taking the course, the students know ARIMA-models and their key properties. In addition, the students are able to model and predict the future behaviour of observed time series using ARIMA-models. The students also know basic concepts of dynamic regression models.

    Note that all the lectures and the exercise classes are given on campus. Remote attendance is not possible.


     --------------------------------------------------------------------------------------------


    Language of Instruction: English, All the lecture materials, exercises, exams etc are in English only!

    Workload: Exercises 24h (2), Lectures 24h, Homework assignments 48 h, reading and studying the lecture materials 36 h

    Assessment Methods and Criteria: Homework assignments, exam

    Study Material: Lecture slides 

    Substitutes for Courses: Mat-2.3128 Prediction and Time Series Analysis

    Prerequisites: MS-A05XX First course in probability and statistics, MS-A02XX Differential and integral calculus 2, and MS-A000X Matrix algebra

    Evaluation: 1-5

    Lectures:  

    Lectures are an important part of this course. Attendance to lectures is not compulsory, but highly recommended. Lecture slides are posted on MyCourses webpage. Please study the slides carefully before solving the corresponding exercises.

    Exercises:

    Exercises are an important part of this course. Attendance to exercises is not compulsory, but highly recommended. You can get exercise points if you solve the homework problems. In order to get the points you have to attend the exercises and be ready to present your solutions in the exercise group. Correct solutions to homework problems are provided in the exercises only.

    The exercise points are valid until the end of June 2023.

    There are several exercise groups. Please attend one of the theory exercise groups and one of the computer exercise groups.

    How to pass this course?

    You are expected to

    Study the lecture slides carefully.

    Participate to exercises and solve your homework problems - not compulsory, but highly recommended - max 12 points.

    Take the exam - max 30 points.

    Max total points = 30 + 12 = 42: You need at least 21 points (or 15 points from the exam) in order to pass the course.


    How to get a good grade?

    Work hard on your homework assignments.

    Be active in the exercises!

    Study for the exam!

    Grading is based on the total points as follows: 

    21p (or 15p from the exam) -> 1,  25p -> 2,  28p -> 3,  31p -> 4,  34p -> 5.