Jan Karasek

Towards an automatic design of non-cryptographic hash function

This paper presents an automatic approach to a non-cryptographic hash function design based on grammar guided genetic programming. The paper describes how it is possible to design a non-cryptographic hash function, implementation issues such as terminal and nonterminal symbols, fitness measure, and used context-free grammar. The main aim of this paper is to link the expert knowledge in the design of non-cryptographic hash function and the process of automatic design which can try many more combinations then an expert can. The hash function automatically designed in the paper is competitive with human design and it is compared with the most used non-cryptographic hashes in the field of speed of processing and in the field of collision resistance. The results are discussed in the last section and further improvement is mentioned.

Evolutionary improved object detector for ultrasound images

Object detection in ultrasound images is difficult problem mainly because of relatively low signal-to-noise ratio. This paper deals with object detection in the noisy ultrasound images using modified version of Viola-Jones object detector. The method describes detection of carotid artery longitudinal section in ultrasound B-mode images. The detector is primarily trained by AdaBoost algorithm and uses a cascade of Haar-like features as a classifier. The main contribution of this paper is a method for detection of carotid artery longitudinal section. This method creates cascade of classifiers automatically using genetic algorithms. We also created post-processing method that marks position of artery in the image. The proposed method was released as open-source software. Resulting detector achieved accuracy 96.29%. When compared to SVM classification enlarged with RANSAC (RANdom SAmple Consensus) method that was used for detection of carotid artery longitudinal section, works our method real-time.

An efficient and lossless fingerprint encryption algorithm using Henon map & Arnold transformation

In this paper two stage biometric data protection scheme is being proposed using permutation and substitution mechanism of the chaotic theory which is lossless in nature. Arnold transformation and Henon map is used to design an efficient encryption system. The encryption method is aimed at generating an encrypted image that will have statistical properties completely dissimilar from the original image analysis which will make it difficult for any intruder to decrypt the image. The performance of the method has been experimentally analyzed using statistical analysis and correlation based methods. Correlation coefficient analysis is done to evaluate the behavior of pixels in horizontal and vertical directions and the results are found to be encouraging. This protection scheme provides the ability to encrypt the data and secure it from unauthorized users. Upon decryption the data is completely recovered making this scheme a lossless and efficient method of biometric data security.

Color image (dis)similarity assessment and grouping based on dominant colors

The computer vision connected to image understanding becomes more and more important in everyday life. This paper concerns the image (dis)similarity assessment and grouping. The main contribution of this paper is the method for image (dis)similarity assessment based on dominant colors. The experimental results showed better results than the Direct Pixel Similarity and Color Histograms and method proved to be capable of detecting images similar to the target image.

Java evolutionary framework based on genetic programming

Automatic optimization techniques, such as evolutionary algorithms, have become popular in the recent years as a general, simple, robust, and scalable solution which can be applied when other optimization method fails. Recently, many evolutionary and/or genetic based optimization frameworks and libraries have been developed and lot of them is freely available. On the other hand, there are not many tools in optimization field that allows the researchers to implement own code, modify existing code or compare different algorithms. This paper proposes a new grammar driven genetic programming based framework implemented in cross-platform Java programming language which allows to implement own code, modify existing, and analyze algorithms. The framework described in this paper addresses the problem of flexibility, modularity, multiplatformness, and presents a general architecture for evolutionary optimization based on genetic programming driven by context free grammar distributed under the LGPL license suitable for both scientific and business applications. In the paper is described a design of the framework, the motivation for development, and two use-cases.

Logistic Warehouse Process Optimization Through Genetic Programming Algorithm

This paper introduces process planning, scheduling and optimization in warehouse environment. The leading companies of the logistics warehouse industry still do not use planning and scheduling by automatic computer methods. Processes are planned and scheduled by an operational manager with detailed knowledge of the problem, processed tasks and commodities, warehouse layout, performance of employees, parameters of equipment etc. This is a quantum of information to be handled by a human and it can be very time-consuming to plan every process and schedule the timetable. The manager is usually also influenced by stress conditions, especially by the time of holidays when everyone is making supplies and the performance of the whole warehouse management goes down. The main contribution of this work is (a) to introduce the novel automatic method for optimization based on the evolutionary method called genetic programming, (b) to give a description of a tested warehouse, and (c) to show the metrics for performance measurement and to give a results which states the baseline for further research.

Genetic programming based classifier in Viola-Jones RapidMiner Image Mining Extension

This paper presents a new approach to the classifier design used in the Viola-Jones object detector implemented in RadpidMiner Image Mining Extension. The new approach to the classifier design proposed in this paper is in fact creation of a classification tree designed by a genetic programming algorithm. The resulting classifier is used as an alternative approach to the standard cascade classifier designed by a genetic algorithm. In this paper, a classifier design is shown, the incorporation into the Viola-Jones operator is described, and experimental results of face classification process are depicted and compared to the standard cascade classifier designed by genetic algorithm.

Optimization of logistic distribution centers process planning and scheduling

This paper describes a novel method for solving the problem of automatic planning and scheduling of work-plans in logistic distribution centers. The solution of the problem is based on well-known scheduling problems such as Job-Shop Scheduling Problems or Vehicle Routing Problems. By the time of writing this article, the key representatives of the logistics and warehousing industry do not use fully automated processes for work scheduling. The purpose of this paper is to connect the scientific result with demands of the companies in logistics and warehousing industry. The main contribution of this paper is a) to describe the motivation for solving the problem of logistic and warehousing companies, b) to describe a set of benchmarks and to give the reference layout of the warehouse, and c) to present a baseline results obtained by a genetic programming.

A novel cardiovascular decision support framework for effective clinical risk assessment

The aim of this study is to help improve the diagnostic and performance capabilities of Rapid Access Chest Pain Clinics (RACPC), by reducing delay and inaccuracies in the cardiovascular risk assessment of patients with chest pain by helping clinicians effectively distinguish acute angina patients from those with other causes of chest pain. Key to our new approach is (1) an intelligent prospective clinical decision support framework for primary and secondary care clinicians, (2) learning from missing/impartial clinical data using Bernoulli mixture models and Expectation Maximisation (EM) techniques, (3) utilisation of state-of-the-art feature section, pattern recognition and data mining techniques for the development of intelligent risk prediction models for cardiovascular patients. The study cohort comprises of 632 patients suspected of cardiac chest pain. A retrospective data analysis of the clinical studies evaluating clinical risk factors for chest pain patients was performed for the development of RACPC specific risk assessment models to distinguish between cardiac and non cardiac chest pain. A comparative analysis case study of machine learning methods was carried out for predicting RACPC clinical outcomes using real patient data acquired from Raigmore Hospital in Inverness, UK. The proposed framework was also validated using the University of Clevelands Heart Disease dataset which contains 76 attributes, but all published experiments refer to using a subset of 14 of them. Experiments with the Cleveland database (based on 18 clinical features of 270 patients) were concentrated on attempting to distinguish the presence of heart disease (values 1, 2, 3, 4) from absence (value 0). The new clinical models, having been evaluated in clinical practice, resulted in very good predictive power, demonstrating general performance improvement over benchmark multivariate statistical classifiers. As part of these case studies, various online RACPC risk assessment prototypes have been developed which are being deployed in the clinical setting (NHS Highland) for clinical trial purposes.

Multi-GPU implementation of k-nearest neighbor algorithm

Using modern Graphic Processing Units (GPUs) becomes very useful for computing complex and time consuming processes. GPUs provide high-performance computation capabilities with a good price. This paper deals with a multi-GPU OpenCL implementation of k-Nearest Neighbor (k-NN) algorithm. The proposed OpenCL algorithm achieves acceleration up to 750x in comparison with a single thread CPU version. The common k-NN was modified to be faster when the lower number of k neighbors is set. The performance of algorithm was verified with two GPUs dual-core NVIDIA GeForce GTX 690 and CPU Intel Core i7 3770 with 4.1 GHz frequency. The results of speed up were measured for one GPU, two GPUs, three and four GPUs. We performed several tests with data sets containing up to 4 million elements with various number of attributes.