Home » Events 2007

Events 2007

14.12.2007: Doctoral defence of Merja Oja

HIIT researcher Merja Oja presents her doctoral thesis Methods for Exploring Genomic Data Sets: Application to Human Endogenous Retroviruses in a public dissertation at the Helsinki University of Technology. The custos of the event is professor Samuel Kaski. The opponent is Professor Hiroshi Mamitsuka, Kyoto University, Japan. [väitöstiedote]

14.12.2007: HIIT seminar: Patrik Hoyer

HIIT seminars in fall 2007 will be held in hall **B222** of Exactum,
on Fridays starting at 10:15 a.m. Coffee available from 10.

Fri Dec 14
Patrik Hoyer

Exploiting both non-gaussianity and conditional independencies for causal discovery

Abstract:

I will describe my recent visit to Carnegie Mellon University in Pittsburgh, both in terms of the general experience and in terms of the scientific results. In particular, I will describe techniques for learning linear causal models that combine the strengths of their methods based on conditional independencies and our algorithms based on non-gaussianity.

This is joint work with Peter Spirtes, Richard Scheines, Clark Glymour, Joseph Ramsey and Gustavo Lacerda at Carnegie Mellon University, Shohei Shimizu at the Institute of Statistical Mathematics in Tokyo, and Aapo Hyvärinen at HIIT.

12.12.2007: HIIT Christmas party

HIIT's Christmas party will be organized on Wednesday 12 December at 5 p.m. The place is Hotel Katajanokka, Vyökatu 1.

7.12.2007: HIIT seminar: Petteri Kaski

HIIT seminars in fall 2007 will be held in hall **B222** of Exactum,
on Fridays starting at 10:15 a.m. Coffee available from 10.

Fri Dec 7

Petteri Kaski
Enumeration of the 1-factorizations of the complete graph K_{14}

Abstract:
In how many different ways can one organize a round-robin football tournament for 14 teams? Two answers conditioned on the precise definition of ``different'' are 98758655816833727741338583040 and 1132835421602062347. This talk provides some background on the problem and discusses the combinatorial algorithms used to compute these numbers.

This is joint work with Patric Östergård.

30.11.2007: HIIT seminar: Jarkko Tikka

HIIT seminars in fall 2007 will be held in hall **B222** of Exactum,
on Fridays starting at 10:15 a.m. Coffee available from 10.

Fri Nov 30
Jarkko Tikka
Two approaches to select relevant inputs for RBF network

Abstract:

In regression problems, making accurate predictions is often the primary goal. Neural networks provide good generalization in many cases, but their interpretability is usually limited. The dependencies between input
and output variables is not clear at all. However, the contributions of input variables in the prediction of output would be valuable information in many real world applications. Two input selection algorithms for Radial Basis Function (RBF) networks are presented to increase the interpretability of the models. Firstly, a sequential input selection algorithm is proposed. It is a backward selection type algorithm based on the partial derivatives of the RBF network. Secondly, a input selection method, which is based on a constrained cost function, is shown. The input selection algorithms are applied to both simulated and benchmark data sets and the results are convincing. That is, both algorithms yield accurate prediction models, which are also parsimonious in terms of input variables.