LEARNING OUTCOMES
Develop understanding and skills for using the tools of probability, signal and systems and learning theory to estimate signals and parameters of interest, to detect events from data, and to learn from data. Develop understanding of multivariate analysis as well. Develop understanding how to identify the optimal estimator/detector or at least bound the performance of any estimator/detector. Develop understanding and skills for using detection approaches (statistical hypotheses testing). Understanding radar detection. Practicing basic concepts such as: sufficient statistics, bias and mean squared error, maximum likelihood, Bayesian estimation and learning.
Credits: 5
Schedule: 10.09.2024 - 26.11.2024
Teacher in charge (valid for whole curriculum period):
Teacher in charge (applies in this implementation): Sergiy Vorobyov, Esa Ollila, Visa Koivunen
Contact information for the course (applies in this implementation):
Lecturers: Profs. Sergiy Vorobyov, Esa Ollila, Visa Koivunen
Teaching assistant: Dr. Kai Dong
You can contact us by email (firstname.lastname@aalto.fi)
CEFR level (valid for whole curriculum period):
Language of instruction and studies (applies in this implementation):
Teaching language: English. Languages of study attainment: English
CONTENT, ASSESSMENT AND WORKLOAD
Content
valid for whole curriculum period:
Statistical modeling and basic distributions. Parameter estimation. Hypothesis testing detection theory. Basics of learning theory.
applies in this implementation
Prerequisites: Basics on Probability; Matrix Calculus, Signals and Systems
Motivation: In every digital system that measures, generates, transmits, processes, and uses data, i.e., in every digital system that you will possibly deal with independent on a specific application, you will need to
1. Estimate signals and parameters of interest
2. Detect events from data
3. Learn and infer from data.
This course will introduce you to the main principles and approaches for addressing such tasks.
Learning Outcomes:
1. Develop understanding and skills for using the tools of probability, signal and systems and learning theory to estimate signals and parameters of interest, to detect events from data, and to learn from data.
2. Starting from the univariate case, develop understanding of multivariate analysis as well. Develop understanding how to identify the optimal estimator/detector or at least bound the performance of any estimator/detector.
3. Develop understanding and skills for using detection approaches (statistical hypotheses testing): Neyman-Pearson test, likelihood ratio test, matched filter detection, sequential test. Understanding radar detection.
4. Basic estimation concepts and the method of maximum likelihood, basic concepts of supervised learning and methods such as linear and quadratic classification.
Assessment Methods and Criteria
valid for whole curriculum period:
Lectures, exercises, assignments, final exam.
applies in this implementation
Grading: All components of the class give approximately the same percentage for the final grade.
Example: with 6 homework assignments and no final exam, each assignment will give on average 16-17% for the final grade - in fact, some initial assignments will give 16% and some later assignments - 17%.
Homework Assignments: It is expected that there will be 6 homework assignments, but we will have to plan it more carefully by talking to you and optimizing the schedule. The plan is to have lighter assignments more frequently!
Exam: we reserve an option to have a light exam
Workload
valid for whole curriculum period:
Lectures, exercises, assignments, final exam, independent.
Attendance in some contact teaching may be compulsory.
applies in this implementation
The work load consists of in class work (lectures and exercise sessions) and home works for solving problems from the assignments at home.
Lectures, exercises, assignments, final exam approximately 50 h, independent work approximately 83 h, total 133 h
Attendance in some contact teaching may be compulsory.
DETAILS
Study Material
applies in this implementation
Lecture slides/notes!!! It is necessary to come to every lecture.
Books:
· Roy D. Yates and David J. Goodman, Probability and Stochastic Processes – A Friendly Introduction for Electrical and Computer Engineers, John Wiley & Sons, Inc., 2nd Ed.
· M. Hayes, Statistical Digital Signal Processing and Modeling, Wiley & Sons, 1996.
· Samuel Stearns and Ruth David, Signal Processing Algorithms in Matlab, Prentice-Hall, 1996.
Substitutes for Courses
valid for whole curriculum period:
Prerequisites
valid for whole curriculum period:
FURTHER INFORMATION
Further Information
valid for whole curriculum period:
Teaching Language: English
Teaching Period: 2024-2025 Autumn I - II
2025-2026 Autumn I - II
Details on the schedule
applies in this implementation
Schedule (tentative, subject to change)
Exercise (Mon 12:15-14:00)
Noether 1572, Kide
Lecture (Tue 9.15-12:00)
T6 A136, Computer Science building
10.9
Lect. 1: Sergiy Vorobyov
17.9
Lect. 2: Sergiy Vorobyov
23.9
Exer. 1: Kai Dong
24.9
Lect. 3: Sergiy Vorobyov
1.10
Lect. 4: Sergiy Vorobyov
7.10
Exer. 2: Kai Dong
8.10
Lect. 5: Sergiy Vorobyov
Exam week (no teaching)
22.10
Lect. 6: Sergiy and Esa
28.10
Exer. 3: Kai Dong
29.10
Lect. 7: Esa Ollila
5.11
Lect. 8: Esa Ollila
11.11
Lect. 9: Visa Koivunen
12.11
Exer. 4: Kai Dong
19.11
Lect. 10: Esa Ollila
25.11
Exer. 5: Kai Dong
26.11
Lect. 11: Esa Ollila