Contact Details

WWW	http://www.unige.ch/~kalousis
e-mail
Phone	+41 22 546 35 84
Postal Address	University of Applied Sciences Western Switzerland
	Department of Business Informatics
	Battiment F,
	7, Route de Drize
	1227 Carouge 4, Switzerland

I do no longer maintain the present page. For more recent information please consult the Data Mining and Machine Learning team page (DMML).

Since 2011 I hold a professor position in the University of Applied Sciences, Western Switzerland, at the Department of Business Informatics where I teach and do research in data mining and machine learning with a small team of researchers. Prior to that I was a reasearch associate within the Computer Science Department of the university of Geneva to which I am still affiliated as a “chargé de cours” , where I collaborate closely with Prof. Stephane Marchand-Maillet VIPER group.  

My research work has two axes. The first deals with the development of novel methods and algorithms in the domains of data mining (DM), and machine learning (ML). The second concerns the use of these tools to solve real world problems in a variety of domains such as biomedical problems, sensor data analysis, chemoinformatics, traffic prediction and more. More often than not the first axis feeds on ideas and challenges set forth by the second.

Research in Data Mining and Machine Learning : Throughout the years I have worked on various aspects such as the development of kernel- and metric-based learning algorithms for problems with complex representational requirements, feature selection and the study of model stability of feature selection and machine learning algorithms, the development of similarity measures for learned models. I have also worked on meta-learning and meta-mining and the automatic composition of knowledge discovery workflows tailored to the requirements of specific users and application problems. The following is a list of research directions I have worked on:

• Learning with complex representations
• Stability of learned models
• Feature weighting and selection
• Meta-learning and meta-mining

a more accurate description of my current research interests is given by my publications list.

Applied Research: The application domain to which I have mostly focused in the past is the biomedical domain and namely the construction of predictive, diagnostic and prognostic models for various pathologies, a significant part of the work concerned the development of analytical tools for mass spectrometry data. You can find here here an obselete description of the different application problems I have worked on. More recently I am looking in forecasting problems applied to areas such as traffic prediction and stream analytics applied on sensor data.

Funded Projects : Throughout the years my research has been supported by a number of EU and Swiss projects; an incomplete list can be found here.

Publications : Manually maintained list, Google Scholar Profile, Business Informatics School paper repository (2011-)

Positions : If you are interested in joining us take a look here for our current open poistions or contact me.