Ilija Tanackov

Golden ratio (sectiona aurea) in markovian ants AI hybrid

The exponential pheromone signal of the AI hybrid of Markovian Ant Queuing System - MAQS, is divided into the spatial and deposit pheromone fractions which have the identical values. A new hybrid is formed. The convolution of two new exponential signals has the Erlang distribution. Introducing the interstate in the process of markovization the Erlang Queuing Ant System - EQAS, is solved. Comparison of the average distance between artificial ants in MAQS and EAQS gave particular numerical specificity. The average distances are in ? equilibrium. Constant ? is a famous constant of the Golden ratio.

The spatial pheromone signal for ant colony optimisation

The effect of the passive insecticide on the ant colony Monomorius pharaonis is localised with minor losses -- only one ant. The information on the insecticide location is transferred through the colony in all directions with great speed. After deserting the basic trail, a rapid consolidation of the new ant colony is probably established by the spatial pheromone signal. A simulation model for the time calculation and the number of ants necessary for the formation of the shortest way between the nest and the fictive food source was formed. The basic ant performances have a prevailing part in the shortest trail formation and those are: the range of the radius pheromone signal and the intensity of the pheromone trail evaporation.

An approach of soft computing applications in clinical neurology

This paper briefly introduces various soft computing techniques and presents miscellaneous applications in clinical neurology domain. The aim is to present the large possibilities of applying soft computing to neurology related problems. Recently published data about use of soft computing in neurology are observed from the literature, surveyed and reviewed. This study detects which methodology or methodologies of soft computing are frequently used together to solve the specific problems of medicine. Recent developments in medicine show that diagnostic expert systems can help physicians make a definitive diagnosis. Automated diagnostic systems are important applications of pattern recognition, aiming at assisting physicians in making diagnostics decisions. Soft computing models have been researched and implemented in neurology for a very long time. This paper presents applications of soft computing models of the cutting edge researches in neurology domain.