Inquiry (E-Journal - Faculty of Business and Administration, International University of Sarajevo)
Not a member yet
209 research outputs found
Sort by
Support Vector Machines for Predicting Protein Structural Classes via Pseudo Images Derived From Amino Acid Sequences
No full textSVM is one of the most widely used and powerful classification algorithms to predict protein structural classes. Via radial base functions, SVM maps the linearly non separable input data to a higher dimensional space where it is almost separable by a hyperplane. First using the training data, six hyperplanes that separate pair wise the four classes training data are constructed. Then the expertise of these six SVMs genuinely aggregated to classify the testing data into four classes. The validation is performed by a boot strap technique. The 33 dimensional data that represents proteins of the data set is derived from pseudo images of proteins that stems from their amino acid sequences. In spite of the simplicity of the features, a Q3 accuracy around 75% is achieved
Fuzzy to Random Uncertainty Alignment
Get PDFThe objective of this paper is to present new and simple mathematical approach to deal with uncertainty alignment between fuzzy and random data. In particular we present a method to describe fuzzy (possibilistic) distribution in terms of a pair (or more) of related random (probabilistic) events, both fixed and variable. Our approach uses basic properties of both fuzzy and random distributions. We show that the data fuzziness can be viewed as a non uniqueness of related random events. We also show how fuzzy-random consistancy principle can be given precise mathemtaical meaning. Various types of fuzzy distributions are examined, special cases considered, and several numerical examples presented
Power Distribution Management System revisited: Single-thread vs. Multithread Performance
Get PDFPower Distribution Management System (PDMS) uses very sophisticated algorithms to deliver reliable and efficient functioning of power distribution networks (PDN). PDNs are represented using very large sparse matrices, whose processing is computationally very demanding. Dividing large PDNs into smaller sub-networks results in smaller sparse matrices, and further processing each sub-network in parallel significantly improves the performance of PDMS. Using multithreading to further process each sub-network however degrades PDMS performance. Single-thread processing of sub-network sparse matrices gives much better performance results, mainly due to the structure of these matrices (indefinite and very sparse) and synchronization overhead involved in multi-thread operations. In this paper an overview of PDMS system is presented, and its performance given single-thread and multiple threads is compared. The results have shown that for some applications, single-threaded implementation in multi-process parallel environment gives better performance than multithreaded implementation
Protein Secondary Structure Prediction by Using PSSM Pseudo Digital Image of Proteins
Get PDFProtein secondary structure prediction is one of the hot topics of bioinformatics and computational biology. In this article we present a new method to predict secondary structure of proteins. PSSMs of proteins are used to generate pseudo image of proteins. These protein images are used to extract digital image features. Digital image features vectors used for similarity analysis. We believe that PSSM pseudo digital images of proteins could help us to represent protein global intrinsic information in order find globally similar proteins and use these similar proteins during prediction. Highest prediction accuracy for Q3 recorded as 72.1% by using the system. Beside the high accuracy, this method allows us to shorten computational time for predicting secondary structure of proteins
Gender Segregation in Employment in Bosnian and Herzegovinian Labor Market
Get PDFThis paper analyses employment segregated market, as method of discrimination that is rather common and present in Bosnia and Herzegovina’s (B&H) labor market. Data on B&H’s labor market were gathered, and examined by using a Mann Whitney test. Since this nonparametric test does not depend on normality of data, it was a best fitting test, for two independent populations, male and female population. Results of our analysis on analysis of person in employment by group of economics activities, shown that in B&H more men are employed than women, and women appear to be more economically inactive in all three groups of sections of economics activities that are selected for the analysis namely agriculture, industry and service sector.
In order to resolve this problem several recommendations were given among which primary research on education attainment and employment opportunities should be conducted on the country level. It is needed to insure that gender equality law and gender action plan will be fully and equally implemented across the country, at all levels and entities
Malicious Web Sites Detection using C4.5 Decision Tree
Get PDFThe technology advancement poses the challenge to the cybercriminals for doing various online criminal acts, such as identity theft, extortion of money or simply, viruses and worms spreading. The common aim of the online criminals is to attract visitors to the Web site, which can be easily accessed by clicking on the URL. Blacklisting seems not to be the successful way of marking Web sites with the “bad” content, considering that many malicious Web sites are not blacklisted. The aim of this paper is to evaluate the ability of C4.5 decision tree classifier in detecting malicious Web sites, based on the features that characterize URLs. The classifier is evaluated through several performance evaluation criteria, namely accuracy, sensitivity, specificity and area under the ROC curve. C4.5 decision tree classifier achieved significant success in malicious Web sites detection, considering all four criteria (accuracy 96.5, sensitivity 96.4, specificity 96.5 and area under the curve 0.958)
Comparison of expectation-maximization clustering and logistic regression on categorization of planets with known properties
Get PDFAnalysis of the exoplanet data is the top priority of astrophysicists today. With the increasing incoming information there is a need for an efficient and reliable algorithm. The data is taken from exoplanet data explorer which was cross checked and filtered with NASA’s known categorization. These were then sorted into 5 categories: Dwarfs, Terrestrial, Icy, Jovian and Giant planets. This paper compares expectation-maximization clustering algorithm as an unsupervised and logistic regression as a supervised machine learning methodologies. Comparatively, logistic regression outperformed EM, indicating it cannot be used to sort through the incoming data. Further analysis is necessary
Intergeneration Social Mobility as a Markov Process
Get PDFIntergeneration social mobility is an old concern in both sociology and economics and refers to a change in the status of family members from one generation to the next. In the line with Markov Chain Theory, in this paper, we provide estimates for intergenerational mobility, which is measured in terms of probabilities. We believe that the social mobility as many other natural and social science process can be represented by Markov matrices.
The results from this study show that after 7 generations the distribution has converged to its stationary point. Meaning, if there is no policy initiative to shift the intergeneration immobility; UK will remain with distribution showing inequality and different opportunities for the young generations depending on their parental background. In addition, over time the number of individual belonging to low-income class has increased, from 0.21 to 0.289. This implies that in UK, the income inequality has been increasing trough the period. The problem is formulated by using the Wolfram Mathematical Programming System
Volatility Estimation Through Historical Prices of Indexes
Get PDFThe volatility is the topic that has been researched in last few decades in various directions. One group of these methods includes the characteristics accumulated through the historical movements of the price of a specific instrument. Other than past data, most of them include other factors such as the stochastic part. In this paper revised are three methods of EWMA, ARCH, and GARCH
Structural Bioinformatics Analysis On Diabetes Associated Proteins
No full textThe Human Diabetes Proteome Project (HDPP) mentioned more than 1000 diabetes associated proteins. 400 diabetes associated proteins whose structure is not analyzed yet are selected from the database. Each proteins structure is analyzed via prediction tools in order to reveal structural similarity with IAPP (Islet Amyloid Polypeptide) whose role is well characterized in diabetes. Through similarity analysis between proteins and in region of disulfide bridge formation, we aimed to find similar possible misfolding patterns in other diabetes associated proteins. Result indicates that Calcitonin Gene Related Peptide I is the only proteins who shows high structural and conformational similarity with IAPP. We believe that one of the possible enrolments of the protein Calcitonin Gene Related Peptide 1 to the diabetes can be similar conformational changes due to disulfide bridge formation break. Finally we propose that the accumulation of misfolded Calcitonin Gene Related Peptide 1 is similar to other protein based conformational diseases. Further structural analyses needs to be performed to confirm these results