Anomaly detection in server metrics with use of one-sided median algorithm

In this paper we consider the problem of anomaly detection over time series metrics data took from one of corporate grade mail service cluster. We propose the algorithm based on one-sided median concept and present some results of experiments showing impact of parameters settings on algorithm performance. In addition we present short description of classes of anomalies discovered in monitored system. Proposed one-sided median based algorithm shows great robustness and good detection rate and can be considered as possible simple production ready solution.

Friedman and Wilcoxon Evaluations Comparing SVM, Bagging, Boosting, K-NN and Decision Tree Classifiers

This paper describes a number of experiments to compare and validate the performance of machine learning classifiers. Creating machine learning models for data with wide varieties has huge applications in predictive modelling across multiple domain of science. This work reviews state of the art techniques in machine learning classifiers methods with several extent of magnitude in statistics and key findings that will be helpful in establishing best methodological practices for class predictions. Comprehensive comparative review analysis with statistical validations for various machine learning algorithm for SVM, Bagging, Boosting, Decision Trees and Nearest Neighborhood algorithm on multiple data sets is carried out. Focus on the statistical analysis of the results using Friedman-Test and Wilcoxon Test as well as other interpretative metrics like classification rate, ROC, F-measure are evaluated to benchmark results.

Application of Moldflow Simulation in Injection Molding of Plastic Tank

This article presents an application of moldflow simulation to optimize the injection molding process of charge air cooler plastic tank. The work shows the advantages of this kind of simulation software and information that it can provide. It also explains how big role today play simulation softwares and how they can improve product and reduce development cost.

Analysis of Performance and Efficiency of Hardware and Software Firewalls

Firewalls are key elements of network security infrastructure. They should guarantee the proper level of security and, at the same time, the satisfying performance in order to not increase the packet delay in the network. In the paper, we present the comparative study on performance and security of a few firewall technologies including hardware, software and virtual solutions. Three important criteria are considered: the maximal throughput of firewall, the introduced delay and the ability to resist Denial of Service attacks. We report results of experiments, present analysis and formulate a few practical conclusions.

ChR: Dynamic Functional Constraints Checking in R

Dynamic typing of R programming language may issue some quality problems in large scale data-science and machine-learning projects for which the language is used. Following our efforts on providing gradual typing library for Clojure we come with a package chR - a library that offers functionality of run-time type-related checks in R. The solution is not only a dynamic type checker, it also helps to systematize thinking about types in the language, at the same time offering high expressivenes and full adherence to functional programming style.

Software Systems Clustering Using Estimation of Distribution Approach

Software clustering is usually used for program understanding. Since the software clustering is a NP-complete problem, a number of Genetic Algorithms (GAs) are proposed for solving this problem. In literature, there are two wellknown GAs for software clustering, namely, Bunch and DAGC, that use the genetic operators such as crossover and mutation to better search the solution space and generating better solutions during genetic algorithm evolutionary process. The major drawbacks of these operators are (1) the difficulty of defining operators, (2) the difficulty of determining the probability rate of these operators, and (3) do not guarantee to maintain building blocks. Estimation of Distribution (EDA) based approaches, by removing crossover and mutation operators and maintaining building blocks, can be used to solve the problems of genetic algorithms. This approach creates the probabilistic models from individuals to generate new population during evolutionary process, aiming to achieve more success in solving the problems. The aim of this paper is to recast EDA for software clustering problems, which can overcome the existing genetic operators’ limitations. For achieving this aim, we propose a new distribution probability function and a new EDA based algorithm for software clustering. To the best knowledge of the authors, EDA has not been investigated to solve the software clustering problem. The proposed EDA has been compared with two well-known genetic algorithms on twelve benchmarks. Experimental results show that the proposed approach provides more accurate results, improves the speed of convergence and provides better stability when compared against existing genetic algorithms such as Bunch and DAGC.

Low-Cost Dynamic Constraint Checking for the JVM

Using formal methods for software verification slowly becomes a standard in the industry. Overall it is a good idea to integrate as many checks as possible with the programming language. This is a major cause of the apparent success of strong typing in software, either performed on the compile time or dynamically, on runtime. Unfortunately, only some of the properties of software may be expressed in the type system of event the most sophisticated programming languages. Many of them must be performed dynamically. This paper presents a flexible library for the dynamically typed, functional programming language running in the JVM environment. This library offers its users a close to zero run-time overhead and strong mathematical background in category theory.

Convergence Analysis of Inverse Iterative Neural Networks with L2 Penalty

The iterative inversion of neural networks has been used in solving problems of adaptive control due to its good performance of information processing. In this paper an iterative inversion neural network with L2 penalty term has been presented trained by using the classical gradient descent method. We mainly focus on the theoretical analysis of this proposed algorithm such as monotonicity of error function, boundedness of input sequences and weak (strong) convergence behavior. For bounded property of inputs, we rigorously proved that the feasible solutions of input are restricted in a measurable field. The weak convergence means that the gradient of error function with respect to input tends to zero as the iterations go to infinity while the strong convergence stands for the iterative sequence of input vectors convergence to a fixed optimal point.

Convergence Analysis of An Improved Extreme Learning Machine Based on Gradient Descent Method

Extreme learning machine (ELM) is an efficient algorithm, but it requires more hidden nodes than the BP algorithms to reach the matched performance. Recently, an efficient learning algorithm, the upper-layer-solution-unaware algorithm (USUA), is proposed for the single-hidden layer feed-forward neural network. It needs less number of hidden nodes and testing time than ELM. In this paper, we mainly give the theoretical analysis for USUA. Theoretical results show that the error function monotonously decreases in the training procedure, the gradient of the error function with respect to weights tends to zero (the weak convergence), and the weight sequence goes to a fixed point (the strong convergence) when the iterations approach positive infinity. An illustrated simulation has been implemented on the MNIST database of handwritten digits which effectively verifies the theoretical results..

Selected Examples of Applications of New Graphics and Animation Technologies

In recent years there has been a marked increase in the competitiveness of some very interesting (user) applications software within the field of computer graphics and animation. This paper presents an analysis of selected examples of the use of graphic applications software designed for professional use within various areas of human activity, and also focusses on the potential for further development of this software. Graphic applications software that makes use of motion capture, performance capture, time-lapse, morphing, Augmented Reality and the use of avatars in human-computer communication has become increasingly popular, cheap and simple.