Natural odors are complex, multidimensional stimuli. Yet, they are learned and recognized by the brain with a great deal of specificity and accuracy. This implies that central olfactory circuits are optimized to encode these complex chemical patterns and to store and recognize their neural representations. What shape this optimization takes remains somewhat mysterious. Recent results from studi...

Current research on machine learning and related algorithms is focused mainly on numeric paradigms. It is, however, widely supposed that intelligent behavior strongly relies on ability to manipulate symbols. In this paper we present on-line versions of self-organizing maps for symbol strings. The underlying key concepts are average and similarity, applied on strings. These concepts are easily d...

Kohonen's Self-Organizing Map (SOM) is one of the most popular arti cial neural network algorithms. Word category maps are SOMs that have been organized according to word similarities, measured by the similarity of the short contexts of the words. Conceptually interrelated words tend to fall into the same or neighboring map nodes. Nodes may thus be viewed as word categories. Although no a prior...

Models are abstractions of observed real world phenomena or processes. A good model captures the essential properties of the modeled phenomena. In the statistical learning paradigm the processes that generate observations are assumed unknown and too complex for analytical modeling, thus the models are trained from more general templates with measured observations. A substantial part of the proc...

one of the complex documents in the real world is city maps. in these kinds of maps, text labels overlap by graphics with having a variety of fonts and styles in different orientations. usually, text and graphic colour is not predefined due to various map publishers. in most city maps, text and graphic lines form a single connected component. moreover, the common regions of text and graphic lin...

In all mammals, much of the neocortex consists of orderly representations or maps of receptor surfaces that are typically topographic at a global level, while being modular at the local level. These representations appear to emerge in development as a result of a few interacting factors, and different aspects of brain maps may be developmentally linked. As a result, evolutionary selection for s...

This paper introduces modifications of Self-Organizing Maps allowing imputation and classification of data containing missing values. The robustness of the proposed modifications is shown using experimental results of a standard data set. A comparison to modified Fuzzy cluster methods [Timm et al., 2002] is presented. Both methods performed better with available case analysis compared to comple...