LKM carries out research in cognitive modelling, machine learning, neural networks, statistics, picture and data mining. Research results concern the modelling of noisy data related to cognitive, medical, biological and other processes. We are developing, testing and applying new approaches and algorithms for modelling from numeric, symbolic and pictorial data, and new approaches to building, evaluation and explanation of models, derived from data. Recent research is related to evaluating the utility of ordinal attributes, generation of semi-artificial data, analysis of big data with map-reduce approach, analysis of graphs, evaluating the reliability of single models’ predictions in classification and regression, evaluating the reliability of clustering, explaining single predictions by arbitrary classification and regression model, text summarization using archetypal analysis, analysing and modelling of sport data, user profiling by mining the web-logs, recommendation systems, applying evolutionary computation to data mining, mining spatial data with multi-level directed graphs, inductive logic programming, heuristic search methods in clickstream mining and mining of data streams. A notable aspect of much of this research is its application to problems in image analysis, medical diagnosis, ecological modelling, marketing and financial modelling.
Laboratory for Cognitive Modeling
University of Ljubljana
Faculty of Computer and Information Science
Večna pot 113