International journal of scientific research in information systems and engineering (IJSRISE) <p>International journal of scientific research in information systems and engineering (IJSRISE) is an open access international peer reviewed multidisciplinary journal that publishes quality studies related to Information Systems, Social Science, Education and Engineering. Not limited to reports of qualitative case studies, meta-analyses, mixed method studies, action researchers, quantitative experiments, surveys. Also, methodological issues and discussions of conceptual. The editors seek to publish articles from a wide variety of academic disciplines and substantive fields; they are looking for clear and significant contributions to the understanding and/or improvement of educational processes and outcomes.</p> en-US (Samson Oluwaseun Fadiya) (Mr Acheme Odeh Okolobia) Fri, 29 Jun 2018 13:16:48 +0000 OJS 60 A NOVEL APPROACH FOR INTERNET OF THINGS BASED INTELLIGENT WEATHER DATA ACQUISITION IN AIRCRAFT <p>Today, every device which is connected to the Internet collects, shares and augments data of each other. This key feature fosters researchers to use the power of sensors while designing intelligent and cognitive systems. One of the widespread usage area of sensors is weather data acquisition because of their place in internet of things. In previous works, it is not found a well-coordinated approach for collecting and sharing weather data in aircraft. In our approach, we propose a novel system which has the capability of self-exploring, self-learning and self-decision under different weather circumstances. An aircraft equipped by powerful sensors is expected to designate its next step according to weather data including deciding whether to fly or flight safety while on air. By applying genetic (evolutionary) algorithm to aircraft sensors, it becomes possible to optimize parameters of remote sensing devices and analyze them in terms of weather data. According to IOT mechanism of our system, it is provided for all electronic devices to be labeled and to be ready for data exchange. Despite the fact that every mechanism works properly in our system, one prominent limitation becomes discontinuous data flow between different sensors. As a result, our system becomes capable of preventing problems while weather data is being shared among different devices.</p> Doğu Sırt, Evren Dağlarlı, Erke Arıbaş ##submission.copyrightStatement## Creative Commons Attribution - NonCommercial - NoDerivs 4.0 Fri, 29 Jun 2018 00:00:00 +0000 A SMART EXPLORATORY SURVEILLANCE AND POSITION TRACKING SYSTEM FOR A HELIUM BASED UNMANNED AIR VEHICLE <p>By growing attention on UAVs (unmanned air vehicle), it has become pervasive to use them in aerial observation and keep track of other aircraft. A 360° high resolution camera placed on helium based UAV captures video streaming in terabytes everyday as well as sensor data. Despite the fact that every aircraft has such kind of vision tracking and database systems today, they all have their own deficiencies. So, the evaluation of such contents are separated from each other in terms of performance. In this work, we point out a new kind of smart system like ADS-B (Automatic Dependent Surveillance - Broadcast) technology for aircraft. With big data approach, we classify aircraft according to data they produce (volume), the speed of production and use (velocity) and the scope of data (variety). Then we propose a novel, smart and user-friendly interface that allows human-controllers to semiconsciously select different aircraft’s positions on air step by step. Also we deploy a secondary unit on helium based UAV to collect necessary data compatible with the cloud. This communicates with the human-controller on ground for further processes and help decide to take action in urgent cases or to share data for machine-to-machine interaction using novel interface.</p> Doğu Sırt, Evren Dağlarlı, Erke Arıbaş ##submission.copyrightStatement## Creative Commons Attribution - NonCommercial - NoDerivs 4.0 Fri, 29 Jun 2018 00:00:00 +0000 ANN ANALYSIS OF TITANIUM ALLOY BEAMS ACCORDING TO BENDING FOR MEDICAL APPLICATIONS <p>In this study, different cross sectional geometries have been compared under bending stress with three different techniques. To determine the optimal design of the designed parts, 6 different cross sectional geometries (Rectangle, Circle, Equilateral Triangle, Diamond, Ellipse and Rounded Rectangle), 101 different cross sectional areas and 3 different materials (titanium alloys) have been used. Length of the beams and applied moments have been assumed constant. Analytical solution, Finite Element Method (FEM) and artificial neural network (ANN) modelling (number of 1818 models) have been performed according to deformation values. At the beginning, analytical solutions and the FEM were compared to each other by using statistical analysis with respect to Root Mean Square (RMS), Absolute Fraction of Variance (R<sup>2</sup>) and Mean Error Percentage in order to confirm the precision of FEM. After the statistical analysis, resulted deformation values were used as data at the ANN modelling. After the ANN analysis same statistical evaluation has been conducted in order to designate the accuracy of the ANN model. As a result, three different techniques have been conducted and a thorough ANN model has been created. Thanks to this model, optimal dimensions of medical devices can be designed according to bending by using this simple and effective method.</p> Hüseyin A. Çetindağ, İhsan Toktaş, Murat T. Özkan, Aysun E. Kılıçarslan, Hatice N. Ünver ##submission.copyrightStatement## Creative Commons Attribution - NonCommercial - NoDerivs 4.0 Fri, 29 Jun 2018 00:00:00 +0000 ANN MODELLING FOR THE BUCKLING ANALYSIS OF MEDICAL TITANIUM ALLOYS <p>Aim of this study is to investigate the cross sectional geometries of beams which have better buckling performance by using three different techniques. At the modeling process length of beams and applied forces have been assumed constant but cross sectional geometries, cross sectional areas and used materials have been changed at each design. These geometries have been selected as circle, rectangle, equilateral triangle, rhombus (diamond), ellipse and rounded rectangle. Moreover three different material and 101 cross sectional areas which increase from 3mm<sup>2</sup> to 8mm<sup>2</sup> have been taken into account for comparison. Firstly, critical stress values have been calculated analytically, then all design points have been analyzed by using finite element method (FEM). Deviation between these two methods has been obtained by using statistical analysis in order to obtain the reliability of the FEM. After the verification of the FEM according to statistical analysis, artificial neural network has been used in order to perform modelling for the FEM results. The deviation between the FEM and the ANN modeling have been calculated by comparing the statistical analysis results of Root Mean Square (RMS), Absolute Fraction of Variance (R<sup>2</sup>) and Mean Error Percentage. At the end of this study, in addition to complete ANN modelling of the design points, three methods have been compared to each other with respect to deviation among the results which were obtained from 1818 design points. Thus, a reliable and simple solution has been created by ANN modelling. This model is easy to apply and it can be used as an alternative of other solution techniques.</p> Murat T. Özkan, İhsan Toktaş, Hüseyin A. Çetindağ, Aysun E. Kılıçarslan, Hatice N. Ünver ##submission.copyrightStatement## Creative Commons Attribution - NonCommercial - NoDerivs 4.0 Fri, 29 Jun 2018 00:00:00 +0000 CHAOS-BASED DATA ENCRYPTION USING ARNOLD’S CAT MAP <p>Continuous Automorphism of the Torus (CAT) is a group of algebraic functions and are typically used in chaos-based encryption applications. Arnold’s CAT Map is one of the known CAT calculations. The most charming property of Arnold’s CAT Map is a number of repetitions of permutations eventually returns the array into the initial state. The number of repetition to return to the initial state is a function of array dimensions and mapping parameter, unquestionably it is chaotic as well. Utilization of Arnold’s CAT Map in encryption is common especially in image encryption and it is generally preferred for its time efficiency when compared with classical block cipher alternatives. Although substitution and permutation are two essential properties of an encryption algorithm, CAT maps are criticized for conducting permutation only. In this study an encryption algorithm that makes use of Arnold’s CAT Map calculation is proposed and its cryptographic properties are presented.</p> Atila Bostan, Murat Karakaya, Gökhan Şengül ##submission.copyrightStatement## Creative Commons Attribution - NonCommercial - NoDerivs 4.0 Fri, 29 Jun 2018 00:00:00 +0000 DETECTION OF PRE-EPILEPTIC SEIZURE BY USING WAVELET DECOMPOSITION AND ARTIFICIAL NEURAL NETWORKS <p>Epilepsy is one of the most frequent disease all over the world and can give valuable information about the functional situation of the brain. This information is clear when EEG signals are collected and analyzed in an appropriate, accurate and reliable way and they make detection of epileptic seizures is possible. It is difficult to interfere with epileptic seizures because they occur suddenly and randomly thus, detection of epilepsy in the early stage is vital for both patients and medical officials. In this study, firstly both 100 interictal (EEG signals which occur between two epileptic seizures) and 100 healthy EEG signals are normalized in order to detect the pre-epileptic seizure. After the normalization process, each EEG signals are decomposed by performing Wavelet Decomposition (WD) and classified with ANN (Artificial Neural Networks). 4-fold cross-validation technique is used to evaluate the classification performance. LogEn entropy is used to extract features from both interictal and healthy EEG signals and classification accuracy is achieved 97,5%.</p> Talha Burak Alakus, Ibrahim Turkoglu ##submission.copyrightStatement## Creative Commons Attribution - NonCommercial - NoDerivs 4.0 Fri, 29 Jun 2018 00:00:00 +0000 EDGE DETECTION BASED ON ANT COLONY OPTIMIZATION IN SAR IMAGES <p>Object/Target detection is difficult to process due to speckles in SAR images, which provide high radiometric and geometric resolution independent of all atmospheric conditions. By using edge detection method which extracts important information in the image, it is possible to obtain higher accuracy and less processing SAR image for target detection by eliminating these speckles. The ant colony algorithm, which is one of the heuristic optimization methods, is an algorithm based on mathematical models of real ant colony behaviors. In image processing area, Ant Colony Optimization (ACO) provides an effective contribution in some methods such as object/target detection in specific images by using edge detection technique. We aim to eliminate the speckles that make difficult for target detection in SAR images by using Edge Detection based on Ant Colony Optimization, which is an effective optimization method.</p> <p>&nbsp;</p> Murat ŞİMŞEK, Volkan ATEŞ, Murat LÜY ##submission.copyrightStatement## Creative Commons Attribution - NonCommercial - NoDerivs 4.0 Fri, 29 Jun 2018 00:00:00 +0000 EXPERIMENTAL AND NUMERICIAL INVESTIGATION OF CEILING MOUNTED RADIANT PANELS HEAT OUTPUT FOR DIFFERENT INLET-OUTET WATER TEMPERATURES <p>The heat output of ceiling mounted radiant panels with dimensions of 1.5m x 1.9m x 0.052m were investigated for inlet-outlet temperatures of 50/30°C, 60/40°C, 70/50°C, 80/60°C and 90/70°C in a temperature controlled space. The experimental results were compared with results of an accredited laboratory, and a fair agreement was obtained. In addition, computational investigations for the radiant panel with the same dimensions were performed, in a space heated with radiant heater with constant surface temperatures in the range of 40°C – 150°C. It was observed that the numerical results are in agreement with the experimental values. The numerical results showed that, radiation as well as natural convection heat transfer mechanisms occur, and the portion of radiation heat transfer is at a rate of approximately 80%. The heat output increases almost linearly with the increase of temperature.</p> Eda Ergur, Serhat Unver, Tamer Calisir, Senol Baskaya, Huseyin Topal ##submission.copyrightStatement## Creative Commons Attribution - NonCommercial - NoDerivs 4.0 Fri, 29 Jun 2018 00:00:00 +0000 IDENTITY VERIFICATION ON PARTIAL FINGERPRINT IMAGES USING GLOBAL FEATURE VECTORS <p>Although there are several commercial applications available in the market, fingerprint identification and verification is still a challenging task in the field of automated computing. Despite the common usage of fingerprint identification systems, current calculation algorithms need to be more developed, since the variations in scanned image significantly effects the success rate. Thus people are frequently requested to re-scan their finger in order to eliminate false-positive verifications. Unquestionably, scanned fingerprints are subject to several external distortions, as is casual scars, regional bruises and partially-blinded scanners. Hence a portion of the local features can not be identified at all. In this study the efficiency of global-image-feature set is tested on partial fingerprint scans. Findings indicate that the success rate of the global feature comparison approach on partial finger print images is %70 higher than that of local feature comparison.</p> Mohammed Alsubaihawi, Atila Bostan, Gökhan Şengül ##submission.copyrightStatement## Creative Commons Attribution - NonCommercial - NoDerivs 4.0 Fri, 29 Jun 2018 00:00:00 +0000 COMPARISON OF CLASSIFICATION ALGORITHMS: A CASE STUDY FOR PHYSICAL ACTIVITY RECOGNITION <p>Mobile applications that are used in the healthcare domain have become more popular in recent years because of their functionality. Additionally, they are easy to access and cheap when compared to applications which are not compatible with mobile devices. Patients who need to lose weight or exercise regularly are willing to use such mobile applications to recognize and track their daily physical activities in an easier and more accurate way. In this study, some of the most popular classification algorithms such as KNN (K-Nearest Neighbors), LDA (Linear Discriminant Analysis) and SVM (Support Vector Machines) are selected and applied to the data set to compare their performances for the recognition of the physical activities; namely, walking, walking quickly and running.</p> Güler Kalem, Çiğdem Turhan ##submission.copyrightStatement## Creative Commons Attribution - NonCommercial - NoDerivs 4.0 Fri, 29 Jun 2018 00:00:00 +0000 TECHNOLOGICAL ADVANCEMENTS IN THE HEALTHCARE DOMAIN <p>Improvements on technological devices have affected many areas of life. In recent years, the usage of mobile phones and devices have increased rapidly with the enhancement of internet accesibility and decline in the prices of internet usage. Promising technological advancements and newly developed applications in the healthcare domain aim to make life easier, which are mainly utilized for personal healthcare, disease management, tracking patient behaviour and providing easier and more flexible ways for communication between physicians and patients. In this study, some of the existing applications that are developed for the healthcare domain are reviewed and their working principles are investigated.</p> Güler Kalem, Çiğdem Turhan ##submission.copyrightStatement## Creative Commons Attribution - NonCommercial - NoDerivs 4.0 Fri, 29 Jun 2018 00:00:00 +0000 DETERMINATION OF STRESS CONCENTRATION FACTOR FOR AN INFINITE 3D SOLID PLATE WITH A DEEP HYPERBOLIC GROOVE AND ANN APPLICATION <p>The stress concentration factor (SCF) is important for design of machine components. Especially designing machine components, some geometrical shapes may occur such as grooves, notches, holes and curves. Working components of machine are deformed. Thus, SCF should be thought when designing the machine equipment. Material selection is so important for stable system design which deals with the stress concentration. Generally, the reasons of SCF depend on the design features. In this research, a deep hyperbolic groove is analyzed in terms of SCF. There are several methods to examine the SCF. These methods can be named such as Numerical Analysis, Finite Element Analysis (FEA), Experimental Analysis, Artificial Neural Network (shortly ANN) Methods etc.&nbsp; One of them is known as ANN model which is used in this research. ANN model is used for determination of SCF for an infinite three-dimensional groove. Groove shape types and dimensions can be varied. This is impossible to do experiment or do numerical analysis for every each design shape. ANN is economical and useful method for determination of SCF. While use the some of data of numerical or experimental results, the other types of grooves SCF can be determined with high reliability. The ANN works with different learning algorithms. Neuron is the main element of ANN. The infinite plate is exposed to torsional loading. Meanwhile groove radius and differences of dimensions between two grooves are changed. Analytical, empirical (Peterson’s) and ANN model results have been obtained, and error calculations have been analyzed with using these results. ANN model demonstrates that, fast and accurate results can be obtained for SCF of hyperbolic groove.&nbsp;</p> Hakan Kaplan, Ihsan Toktas, Murat Tolga Ozkan ##submission.copyrightStatement## Creative Commons Attribution - NonCommercial - NoDerivs 4.0 Fri, 29 Jun 2018 00:00:00 +0000 EXPERIMENTAL INVESTIGATION OF PHYSICAL PROPERTIES OF THE BRICKS PRODUCED IN KONYA <p>The demand of becoming powerful against environmental conditions caused people to perform many various researches on this subject. Thereby, the combination of clay with water and reacting with fire constituted the formation of fired brick. At first, the human beings have used fired clay to produce pot, bowl, etc. and presented serious advances in such productions. As a result of observations and experiences on fired brick, today's brick production was formed.</p> <p>Besides dividing the spaces in the buildings, the brick walls have some other functions for reinforced concrete and masonry structures. For reinforced concrete buildings, the brick walls have functionality in terms of decreasing the oscillation period of the building during an earthquake, reducing the goods and life losses during a possible earthquake and protecting the building from atmospheric and external harmful effects. For masonry buildings, the brick walls form the load bearing system of the buildings besides the aforementioned functions. In order to fulfill these functions properly, the appropriateness of the physical and mechanical properties of the clay units forming the brick walls to the related standards' criteria has great importance.</p> <p>In this study, were used&nbsp; the bricks produced in Konya&nbsp;&nbsp;&nbsp; belonging to four different companies . The nominal size of the bricks 8,5x19x19, 13,5x19x19, 13,5x19x29 cm. For each brick type by applying physical tests on the specimens,&nbsp; the obtained results were compared with the criteria prescribed in the related standards.</p> Hicran AÇIKEL, Durmuş Ali AÇIKEL ##submission.copyrightStatement## Creative Commons Attribution - NonCommercial - NoDerivs 4.0 Fri, 29 Jun 2018 00:00:00 +0000 ON THE ACCURACY OF TDOA-BASED PASSIVE LOCALIZATION OF EMITTERS <p>This study addresses a problem of passive emitter localization based on time difference of arrivals (TDOAs) in two dimensions using three sensors. For this purpose, firstly, TDOA-based methods that have been proposed in the literature for estimating emitter location are investigated. Among these studies, a method that gives a simple but deterministic solution is determined. As the derivation of the solution is deterministic, developments are required to ensure its usability in practice. This is because, in practice, especially in an operational environment in electronic warfare, exact positions of the sensors may be unknown, and there may be error in TDOA measurements. Obviously, any errors that may be in sensor positions or in TDOA measurements adversely affect the accuracy of passive emitter localization. For this reason, this study aims to analyze the accuracy of the localization method by introducing measurement error on the TDOAs. To this end, simulations representing realistic scenarios in under various conditions are conducted. In this way, measurement error in TDOAs is considered to test, numerically, the limits of proposed localization technique for realistic scenarios. Future works considered to increase the accuracy of the method are also discussed.</p> Yaser Dalveren, Ali Kara ##submission.copyrightStatement## Creative Commons Attribution - NonCommercial - NoDerivs 4.0 Fri, 29 Jun 2018 00:00:00 +0000 OPTIMAL NUMBER OF CLUSTERS <p>Clustering is an important analysis methods in Data Analytics and Pattern Recognition. The process divides the data into groups without any supervision or external labels and it is a subjective analysis as the definition of a cluster is context dependent. Because of this reason many algorithms, like k-nearest neighbors, require the number of clusters to be fixed a priori. Each clustering algorithm depends on a distance metric to identify different groups in the data. Once the number of centers are fixed, each algorithm tries to find the best separation according of its distance measure by using an optimization algorithm. The distance metric determines the shape of the clusters generated. There are algorithms, like Ward, to determine how many clusters we have in a data set and these algorithms also depend on the same distance metrics where many metrics, like Euclidean and its derivatives, generate hyper ellipsoidal clusters and fail in nonlinearly clustered data. Another computationally expensive approach is to run a specific algorithm for different number of cluster centers and try to choose the best number. In this paper, we attempt to analyze the number of clusters using a previously developed Information Theoretical metric called CEF which; in its original use; can separate nonlinear clusters. Data points that are more similar to each other are incrementally joined together using a distance measure to create subclusters like joined data points against the rest of the data. The operation continues until all data elements are consumed. The CEF metric is used to measure the distance between obtained clusters where peaks in the measure indicates strong cluster separation. The method is tested in several artificial and real data sets and its advantages and disadvantages are discussed.</p> Erhan Gokcay, Murat Karakaya Karakaya, Gokhan Sengul ##submission.copyrightStatement## Creative Commons Attribution - NonCommercial - NoDerivs 4.0 Fri, 29 Jun 2018 00:00:00 +0000 PREDICTING STUDENT’S PASS/FAIL STATUS IN AN ACADEMIC COURSE USING DEEP LEARNING: A CASE STUDY <p>In this paper, we propose new classification models based on Deep Learning (DL) to predict a student’s pass/fail status who took the “Probability and Statistics” course at the Computer Engineering Department of Çukurova University. The dataset used to create the models consisted of data related to 132 students and included various variables such as personal information of the students, different quiz and exam scores, conference attendance and overall absence from the course. For comparison purposes, classification models based on three further machine learning classifiers including Multilayer Perceptron (MLP), Cascade Correlation Network (CCN) and Support Vector Machine (SVM) have also been developed. The results show that the DL-based models, in general, exhibit the most successful classification accuracies, ranging from 65.38% to 100.00%. Furthermore, it is seen that inclusion of average quiz, midterm and final exam scores in models have the most improving effect in predicting a student’s pass/fail status in the course.</p> Fatih Abut, M. Can Yüksel, M. Fatih Akay, Shahaboddin Daneshvar ##submission.copyrightStatement## Creative Commons Attribution - NonCommercial - NoDerivs 4.0 Fri, 29 Jun 2018 00:00:00 +0000 A SURVEY OF SPEECH SEPERATION BY DEEP LEARNING <p>In the last decade, image and voice based applications have begun to take an important part in daily life and widely used for various purposes such as recognition, tracking, security, etc. . The most used methods in these fields are based on machine learning. For example conventional neural networks have been applied in many studies to improve accuracy in these fields. Parallel to this progress in learning algorithms, new processor which supports applying learning algorithm on big data and matrix based operations has been developed. The last step in parallel processing is applying deep learning in image and voice applications. From view point of hardware implementation, GPU processors support the learning and testing deep learning algorithms. In the last decade exploiting the capacity of GPU for mathematical operations provided a hardware with high performance and low cost for big matrix calculations. Combining this with novel machine learning techniques made possible to deal with big data&nbsp; and emerge of Deep Learning concept.</p> <p>In order to separate an individual voice from the other in the noisy environment, the proposed study investigates and survey the last studies to find an accurate acceptable step&nbsp; for solving this problem using deep learning and suggests&nbsp; a deterministic approach for similar studies with conventional methods.</p> Emre Oner Tartan, Ali Berkol, Ali Yücelen ##submission.copyrightStatement## Creative Commons Attribution - NonCommercial - NoDerivs 4.0 Fri, 29 Jun 2018 00:00:00 +0000