Estimation, Detection and Classification

Estimation, detection and classification are at the heart of most signal processing systems which are central in Information and Communication Technology (ICT) and also provide the foundation for data analytics in a broader context (e.g. finance, medicine, industry, earth science). This course gives an introduction to the basic techniques for estimation, detection and classification with focus on ICT applications and signal processing aspects. The generality of the tools is shown through a variety of selected problems, using real-world data from biomedical, multimedia, and speech applications. The course is divided into three modules: --- a) Estimation : Introduction, Minimum Variance Unbiased Estimators and Cramer-Rao Lower Bound, Linear Models and Estimators and Least Squares, Maximum Likelihood and Bayesian Estimation --- b) Detection : Introduction, Statistical decision theory, Binary hypothesis, Likelihood ratio test, Bayes risk, Neyman-Pearson, ROC/DET. Detection of respectively deterministic and random signals. --- c) Classification : Introduction, The theoretical optimal classifier, three basic classifier types, estimation and clustering in classifier design, evaluation of performance, short on state-of-the-art including machine learning.

Knowledge : The candidate have (i) principal understanding of the concepts of estimation, detection and classification; (ii) detailed knowledge of the basic methods within the three topic as a gateway to more advanced techniques; (iii) practical understanding of how to model and solve a variety of real-world problems; and (iv) knowledge of the importance of data for design and evaluation. Skills : The candidate can (i) identify the needs for estimation, detection and classification in practical problems; (ii) select appropriate methods for a given problem; (iii) expand autonomously the knowledge with more advanced techniques if necessary; (iv) implement the method in Matlab/Python, and (v) evaluate the quality of a chosen method for a given problem.

Lectures with focus on practical examples, exercises and a group based project.