Ph.D at Abo Akademi University in Finland. Teaching & research positions in Sweden: at Lulea University of Technology, Linkoping University, KTH Royal Institute of Technology (current position). Visiting researcher at University of North Carolina at Chapel Hill, Louisiana State University at Baton Rouge, UTIA of Czech Academy of Sciences, University of Warsaw, Institute of Advanced Study at Aalto University, Visiting lecturer and researcher at University of Helsinki, University of Turku, University of Makerere. Author (co-author) of three monographs.
Has worked in signal processing, biomathematics and genetics groups.
1. Computational skill in probability: Axioms of Probability; Distributions of probability theory; Conditional probability and Expectation; Martingales; Characteristic functions; Multivariate Gaussian; Probability generating functions; Convergence concepts of probability theory; Law of large numbers; Central limit theorem; Stationary stochastic processes; Wiener process; Ornstein-Uhlenbeck Process; Poisson process.
2. Application of A in theory of statistical inference and machine Learning: PAC-theory; Expected risk minimisation; Exponential families of distributions; Bayesian theory; Maximum Likelihood; Bias-Variance Decomposition of data science; Confidence intervals; Model choice in machine learning; Expectation maximisation algorithm; High-dimensional data.
SKILLS:
- Statistics
- Applied Mathematics
- Matlab
- Machine Learning
DATE: 18 Feb - 8 Mar, 2019
DURATION: 3 Week
LECTURES: 3 Hours per day
LANGUAGE: English
LOCATION: Barcelona, Harbour.Space Campus
COURSE TYPE: Offline
WHAT YOU WILL LEARN
ABOUT TIMO
HARBOUR.SPACE
The aim of the course is to introduce basic theories and methods of pure probability theory at an intermediate level and their applications to chosen topics of theory of statistical inference and machine learning. No knowledge of measure and integration theory is required, and only bare first statements of that will be included in the course.
Techniques developed in this course are important in AI, time series analysis, financial analysis, signal processing, econometrics, and other branches of engineering and science. The course gives also a background and motivation for studies of advanced courses in probability and statistics.
TIMO KOSKI
HARBOUR.SPACE UNIVERSITY
DATE: 18 Feb - 8 Mar, 2019
DURATION: 3 Weeks
LECTURES: 3 Hours per day
LANGUAGE: English
LOCATION: Barcelona, Harbour.Space Campus
COURSE TYPE: Offline
All rights reserved. 2017
COURSE OUTLINE
Session 1
Sigma-fields,
Probability space.
Axioms of probability calculus
Session 2
Some Theorems of Probability calculus.
Probability on propositional logic.
PAC-theory of machine learning
Session 3
Distribution functions.
Multivariate random variables
Session 4
Multivariate random variables.
Marginal density, Independence,
Density of a transformed random vector,
Exponential families of distributions
THEORY OF PROBABILITY AND STATISTICS
THEORY OF PROBABILITY AND STATISTICS
BIBLIOGRAPHY