Mariusz Boryczka

Ant Colony Programming for Approximation Problems

A method of automatic programming, called genetic programming, assumes that the desired program is found by using a genetic algorithm. We propose an idea of ant colony programming in which instead of a genetic algorithm an ant colony algorithm is applied to search for the program. The test results demonstrate that the proposed idea can be used with success to solve the approximation problems.

Discretization of Continuous Attributes on Decision System in Mitochondrial Encephalomyopathies

In this work we check how the automatic discretization algorithms generate decision rules for the concrete medical problem - diagnosing mitochondrial encephalomyopathies (MEM). We describe several algorithms for discretization - local and global - of continuous attributes obtained in the second stage of diagnosing MEM. All of these algorithms act together with the data analysis method based on the rough sets theory. This work compares results -- quality of classification rules -- which were obtained using different discretization methods of the continuous attributes.

Ant colony programming with the candidate list

This work develops an idea of the ant colony programming in which an ant colony algorithm is applied to search for the computer program. We shows, that the candidate list (introduced in ant colony system) allows to reduce the algorithms execution time and improves the quality of results.

Multi-cast ant colony system for the bus routing problem

MCACS-BRP, a new Ant Colony Optimization (ACO) based approach to solve the Bus Routing Problem is presented. MCACS is an extension of ACO, where two hierarchically connected casts of ants optimize two different objective functions. In MCACS-BRP, ants collaborate using information about the best results obtained in the particular cast. Experiments with real data from the Municipal Public Transport Union of the Upper Silesian Industrial District (KZK GOP) show that MCACS-BRP is worth further experiments and extensions.

Ant Colony Optimization for the Multi-criteria Vehicle Navigation Problem

The chapter describes the multi-agent ant-based vehicle navigation algorithms, which find multi-criteria optimal route between two points on the map. Presented are various versions of the algorithm, sequential and parallel, including GPU one. Various experiments performed on data of different size show the ability of presented algorithms to find good (near optimal) solutions for large real map. In turn, the parallel AVN algorithm is able to produce even better results in shorter time. Finally, we show that the presented approach may be adapted to run on GPUs and the algorithm’s performance scales very well with growing number of multiprocessors.

Dynamic Parameters in GP and LGP

Genetic Programming (GP) is one of Evolutionary Algorithms. There are many theories concerning setting values of main parameters that determine how many individuals will crossover or mutate. In this article we present a method of building dynamic parameter that will improve fitness function. In this way we create hybrid parameters that affect on individual. For testing we use our own dedicated platform. Our investigations of the best range of each parameter we based on our preliminary experiments.

Genetic Algorithm in Stream Cipher Cryptanalysis

Cryptography nowadays is a very important field of protecting information from falling into wrong hands. One of modern cryptography branch is stream cipher cryptography. This paper focuses on cryptanalysis of such ciphers using genetic algorithm. Genetic algorithm as one of optimisation methods isn’t quite obvious to use in the field of cryptography, nevertheless it can give interesting results. In this article authors look for the shortest equivalent linear system which approximate given keystream with linear shift feedback register.

Ant colony optimization for the pareto front approximation in vehicle navigation

This paper presents an example of a multi-criteria optimization problem for vehicle navigation in the presence of multiple criteria and method of employing Ant Colony Optimization metaheuristic to solve it. The paper presents an approach based on the concept of Pareto optimality and approximation of set of non dominated solutions forming the so-called Pareto front.

Evolutionary and Aggressive Sampling for Pattern Revelation and Precognition in Building Energy Managing System with Nature-Based Methods for Energy Optimization

This chapter presents a discussion on an alternative attempt to manage the grids that are in intelligent buildings such as central heating, heat recovery ventilation or air conditioning for energy cost minimization . It includes a review and explanation of the existing methodology and smart management system . A suggested matrix-like grid that includes methods for achieving the expected minimization goals is also presented. Common techniques are limited to central management using fuzzy-logic drivers, but referred redefining of the model is used to achieve the best possible solution with a surplus of extra energy. Ordinary grids do not permit significant development in the present state. A modified structure enhanced with a matrix-like grid is one way to eliminate basic faults of ordinary grids model, but such an intricate grid can result in sub-optimal resource usage and excessive costs. The expected solution is a challenge for different Ant Colony Optimization (ACO) techniques with an evolutionary or aggressive approach taken into consideration. Different opportunities create many latent patterns to recover, evaluate and rate. Increasing building structure can surpass a point of complexity, which would limit the creation of an optimal grid pattern in real time using the conventional methods. It is extremely important to formulate more aggressive ways to find an approximation of the optimal pattern within an acceptable time frame.

The Role of Randomness in Self-aggregative AntTree Approach

In some research works concerning biomimicry and data mining, new bio-inspired clustering algorithm has been proposed to deal with the difficult problem of a partitioning the data. In this work, a role of randomness in AntTree-based approach is discussed in clustering application. This proposition integrates the random mechanism of inserting ants in the tree representation of partitioning and the concept of attraction of the specific connections in the analyzed structure. In the same time, the role of shoving (dynamically changed) by the dissimilarity between objects has been analyzed. The comparative study concerning ant-based algorithm and the standard DBSCAN approach shows that this proposal achieves results comparable to the best classical approach’s results. This approach shows that randomness improves the results in clustering offered by the AntTree algorithm.