Machine Learning Coffee seminar "Learning of Ultra High-Dimensional Potts Models for Bacterial Population Genomics"

Lecturer : 
Jukka Corander
Event type: 
HIIT seminar
Doctoral dissertation
Respondent: 
Opponent: 
Custos: 
Event time: 
2017-11-13 09:15 to 10:00
Place: 
Exactum D122, Kumpula
Description: 

Jukka Corander, Professor of Statistics, University of Helsinki and University of Oslo

Learning of Ultra High-Dimensional Potts Models for Bacterial Population Genomics

Abstract: The potential for genome-wide modeling of epistasis has recently surfaced given the possibility of sequencing densely sampled populations and the emerging families of statistical interaction models. Direct coupling analysis (DCA) has earlier been shown to yield valuable predictions for single protein structures, and has recently been extended to genome-wide analysis of bacteria, identifying novel interactions in the co-evolution between resistance, virulence and core genome elements. However, earlier computational DCA methods have not been scalable to enable model fitting simultaneously to 10000-100000 polymorphisms, representing the amount of core genomic variation observed in analyses of many bacterial species. Here we introduce a novel inference method (SuperDCA) which employs a new scoring principle, efficient parallelization, optimization and filtering on phylogenetic information to achieve scalability for up to 100000 polymorphisms. Using two large population samples of Streptococcus pneumoniae, we demonstrate the ability of SuperDCA to make additional significant biological findings about this major human pathogen. We also show that our method can uncover signals of selection that are not detectable by genome-wide association analysis, even though our analysis does not require phenotypic measurements. SuperDCA thus holds considerable potential in building understanding about numerous organisms at a systems biological level.

Machine Learning Coffee seminars are weekly seminars held jointly by the Aalto University and the University of Helsinki. The seminars aim to gather people from different fields of science with interest in machine learning. Talks will begin at 9:15 am and porridge and coffee will be served from 9:00 am.

Next talks:

Nov 20, Otaniemi: Perttu Hämäläinen "Towards Intelligent Exergames"

Nov 27, Kumpula: Erik Aurell

Welcome!

 

Machine Learning Coffee seminar "Learning Markov Equivalence Classes of Directed Acyclic Graphs: an Objective Bayes Approach"

Lecturer : 
Guido Consonni
Event type: 
HIIT seminar
Doctoral dissertation
Respondent: 
Opponent: 
Custos: 
Event time: 
2017-10-30 09:15 to 10:00
Place: 
Exactum D122, Kumpula
Description: 

Guido Consonni, Professor of Statistics, Universita Cattolica del Sacro Cuore

Learning Markov Equivalence Classes of Directed Acyclic Graphs: an Objective Bayes Approach

Abstract: A Markov equivalence class contains all the Directed Acyclic Graphs (DAGs) encoding the same conditional independencies, and is represented by a Completed Partially Directed DAG (CPDAG), also named Essential Graph (EG). We approach the problem of model selection among noncausal sparse Gaussian DAGs by directly scoring EGs, using an objective Bayes method. Specifically, we construct objective priors for model selection based on the Fractional Bayes Factor, leading to a closed form expression for the marginal likelihood of an EG. Next we propose an MCMC strategy to explore the space of EGs, possibly accounting for sparsity constraints, and illustrate the performance of our method on simulation studies, as well as on a real dataset. Our method is fully Bayesian and thus provides a coherent quantification of inferential uncertainty, requires minimal prior specification, and shows to be competitive in learning the structure of the data-generating EG when compared to alternative state-of-the-art algorithms.

Machine Learning Coffee seminars are weekly seminars held jointly by the Aalto University and the University of Helsinki. The seminars aim to gather people from different fields of science with interest in machine learning. Talks will begin at 9:15 am and porridge and coffee will be served from 9:00 am.

Next talks:

Nov 6, Otaniemi: Pekka Marttinen "Efficient and accurate approximate Bayesian computation"

Nov 13, Kumpula: Jukka Corander "Learning of Ultra High-Dimensional Potts Models for Bacterial Population Genomics"

Nov 20, Otaniemi: Perttu Hämäläinen "Towards Intelligent Exergames"

Welcome!

Helsinki Algorithms Seminar: "Delegatable and error-tolerant algorithms" Petteri Kaski, Aalto

Event type: 
HIIT seminar
Event time: 
2017-10-19 16:15 to 17:00
Place: 
Exactum B222, Gustaf Hällströmin katu 2B, Helsinki
Description: 

Speaker: Petteri Kaski
Assistant Professor, Aalto University

Delegatable and error-tolerant algorithms

Abstract:

Is it possible to delegate a computation to an unreliable and more powerful counterparty? Can we design algorithms in such a way that not only can their execution be delegated, but a controlled number of adversarial errors can take place during the execution and yet one can recover the desired result? This talk will review theory and engineering efforts to bring such algorithm designs to the computing practice, including some of our recent work.

**

Helsinki Algorithms Seminar is a weekly meeting of researchers in the Helsinki area interested in the art of algorithms and algorithm design, broadly interpreted to cover both theoretical ideas and algorithm engineering on concrete computing platforms. In most cases we have a presentation prepared for each meeting to communicate an idea, a recent result, work-in-progress, or demo, but this should not be at the expense of discussion and simply having fun with algorithms.

Our affiliations are with Aalto University and the University of Helsinki, and accordingly our activities alternate between the Otaniemi Campus of Aalto University and the Kumpula Campus of University of Helsinki, catalyzed by the Helsinki Institute for Information Technology HIIT, under the Algorithmic Data Analysis (ADA) programme.

Welcome!

Machine Learning Coffee seminar "Machine learning for Materials Research"

Lecturer : 
Flyura Djurabekova
Event type: 
HIIT seminar
Event time: 
2017-10-02 09:15 to 10:00
Place: 
Exactum D122, Kumpula
Description: 

Flyura Djurabekova, Department of Physics, University of Helsinki

Machine learning for Materials Research

Abstract: In materials research, we have learnt to predict the evolution of microstructure starting with the atomic level processes. We know about defects -- point and extended, -- and we know that these can be crucial for the final structural (and related mechanical and electrical) properties. Often simple macroscopic differential equations, which are used for the purpose, fail to predict simple changes in materials. Many questions remain unanswered. Why a ductile material suddenly becomes brittle? Why a strong concrete bridge suddenly cracks and eventually collapses after serving for tens of years? Why the wall of high quality steels in fission reactors suddenly crack? Or, why the clean smooth surface roughens under applied electric fields? All these questions can be answered, if one peeks in to atom's behavior imagining it jumping inside the material. But how the atoms "choose" where to jump amongst the numerous possibilities in complex metals? Tedious parameterization can help to deal with the problem, but machine learning can provide a better and more elegant solution to this problem. 

In my presentation, I will explain the problem at hand and show a few examples of former and current application of Neural Network for calculating the barriers for atomic jumps with the analysis of how well the applied NN worked.

Machine Learning Coffee seminars are weekly seminars held jointly by the Aalto University and the University of Helsinki. The seminars aim to gather people from different fields of science with interest in machine learning. Talks will begin at 9:15 am and porridge and coffee will be served from 9:00 am.

Next talks:

October 9, Otaniemi: Daniel Simpson
October 16, Kumpula: Elja Arjas
October 23, Otaniemi: Aristides Gionis
October 30, Kumpula: Guido Consonni

Welcome!

Rajapinta Unconference

Lecturer : 
Event type: 
Conference
Doctoral dissertation
Respondent: 
Opponent: 
Custos: 
Event time: 
2017-11-02 09:00 to 2017-11-03 17:00
Place: 
Dipoli, (Otakaari 24)
Description: 

The first annual Rajapinta unconference will be organized in Dipoli, Otakaari 24. The event is open to all interested in the study of digital society or digital methods for social sciences. Potential themes include using machine learning for social sciences, the design of digital systems used by public government, sociology motivated studies of software and developers and any other cross-disciplinary themes involving social sciences and computing.  Enroll by 10.10, it is free of charge thanks to generous support by HIIT.

Thursday 2.11. is a workshop day has two themes: online research ethics and infrastructures of digital data collection. Click here for more info.

Friday 3.11.is an unconference day, which builds upon the ideas and proposals of the participants. Please see guidelines for making your own proposal here.

The unconference is supported by HIIT and Kone Foundation.

Pages