PEXESO Evolutionary Methods Library is the library of Evolutionary Optimization Methods for Real Domains. It is based on the original Object Oriented Algorithmic model that consists of the multi- operators technology (currently it supports 13 operator types) and "open policy" on the selection strategy (currently 4 selection strategy types). Using this method you have a possibility to compose your own optimization method using some combination of operators and selection strategies, or you can use one of 3 precomposed algorithms. It is provided with several examples and comprehensive HTML documentation.
BBalc is an open framework for the BloodBowl game. The purpose of this library is to provide chess-computer like functions like analyzing the game or generating BloodBowl actions. It provides a versatile interface to access and modify game data. The modular design allows users to extend existing analyse modules to their needs, build new modules from scratch, or even build complete applications based on the BBalc framework.
Tspell is a library and applications for solving Turkish Natural Language Processing (NLP) related computational problems. Turkish, by nature, has a very different morphological and grammatical structure than Indo-European languages such as English. Since it is an agglutinative language like Finnish, even making a simple spell checker is very challenging. Some target problems are: a spell checker, a word analyzer that determines roots and suffixes, a word constructor based on suffixes, and much more.
QSMM, the "QSMM State Machine Model", is a framework for development of non-deterministic intelligent state models and systems with spur-driven behavior. It includes low-level functions for generating optimal actions by the system and high-level functions for building multinode models. In a multinode model, nodes represent components of a system you develop which choose optimal actions using the framework and can correspond to entities external to the system and which behavior is to be learnt. A node can choose optimal actions based on a current node state which is either set manually by your program or is identified automatically by the framework. Probability profiles for a state transition matrix and an action emission matrix of the node can be specified using an assembler program with a user-defined instruction set.
Stochastic discrimination is a general methodology for constructing classifiers appropriate for pattern recognition. It is based on combining arbitrary numbers of very weak components, which are usually generated by some pseudorandom process, and it has the property that the very complex and accurate classifiers produced in this way retain the ability, characteristic of their weak component pieces, to generalize to new data as complexity increases. These utilities provide an implementation of this algorithm.