International Journal of Computer (IJC - Global Society of Scientific Research and Researchers, GSSRR)
Not a member yet
459 research outputs found
Sort by
Swarm Intelligence Based Feature Selection for High Dimensional Classification: A Literature Survey
Get PDFFeature selection is an important and challenging task in machine learning and data mining techniques to avoid the curse of dimensionality and maximize the classification accuracy. Moreover, feature selection helps to reduce computational complexity of learning algorithm, improve prediction performance, better data understanding and reduce data storage space. Swarm intelligence based feature selection approach enables to find an optimal feature subset from an extremely large dimensionality of features for building the most accurate classifier model. There is still a type of researches that is not done yet in data mining. In this paper, the utilization of swarm intelligence algorithms for feature selection process in high dimensional data focusing on medical data classification is form the subject matter. The results shows that swarm intelligence algorithms reviewed based on state-of-the-art literature have a promising capability that can be applied in feature selections techniques. The significance of this work is to present the comparison and various alternatives of swarm algorithms to be applied in feature selections for high dimensional classification
Higher Education Institutions Data Viz 1.0: Visualization Tool for Decision-Making
Get PDFAs part of the Commission on Higher Education’s (CHED) thrust for improving efficiency and effectiveness by simplifying the collection process for all the stakeholders, the CHED DataViz 1.0 tool will drastically improve the availability of data for making informed decisions and efficient generation of reports by presenting it in a visualized format. This research outlines opportunities associated with the implementation and governance of Big Data in higher education through development and implementation of CHED Higher Education Institution DataViz 1.0, a data analytics tool for decision-making
A Comparative Study of Classification Rule Discovery with Ant Colony Optimization: AntMiner
Get PDFRule based classification is the fundamental and important task of data classification. To discover classification rules, ant colony optimization algorithms are successfully applied that follow a sequential covering approach to build a list of rules. AntMiner Rule Based Classification algorithms are inspired from self- organizing behaviour of ant colonies. In this paper, we presented a study on Ant Colony Optimization Algorithm, AntMiner, c_AntMiner, c_AntMiner2, c_AntMiner PB and conducted experiments to find predictive accuracy against well-known rule induction algorithms JRIP and PART and results shows that AntMiner and its variants shows comparable as well as better performance in some datasets taken in the experimental study
Critical Review in Computer Science: Identification of the Relationship of the Social and Human Factors Related to Teamwork with Software Development Team’s Productivity
Get PDFThis research aims to explore crucial matters related to software development team collaboration that affects productivity. The collaboration component will be discussed in this research including attitude competencies, skill competencies, and knowledge competencies. The results of this study indicate that the components and relationships and human factors influence the productivity of the software development team
New Normal and Abnormal Red Blood Cells Features for Improved Classification
Get PDFThis paper focused obtaining new features for improved classification of red blood cells (RBCs). RBCs varies according to shapes, colors and sizes. Abnormal RBCs may be caused by anemia. Abnormal RBCs has great similarities among each other causing difficulties in medical diagnosis. In this work, spatial, spectral statistical features and geometrical features of RBCs are extracted from 1000 normal and abnormal RBCs. The extracted features are reduced using Principal Component Analysis (PCA) and tested with different types of machine learning algorithms for classification. Classifications were evaluated for high sensitivity, specificity, and kappa statistical parameters. The classifications yielded accuracy rates of 97.9%, 98% and 98% for discriminative (SVM), generative (RBFNN) and clustering (K-NN) algorithm respectively, which is an improvement over previous works
A Comprehensive Study of Optimal Linear Pre-coding Schemes for a Massive Mu-MIMO Downlink System; a Survey
Get PDFMassive Multi-User Multiple input Multiple Output (MU-MIMO) has become one of the leading area in terms of research in wireless communication due to the fact that the number of users and applications have increased tremendously, among all the aspects of massive mu-mimo systems out there, this manuscript focuses on linear precoding for downlink (DL) system at the base station(BS). This manuscript provides a comprehensive survey of precoding techniques for downlink transmission under a single-cell (SC) scenario. In a single-cell (SC) scenario the performance of the precoding techniques, Zero-Forcing (ZF), Match Filter (MF),Truncated polynomial Expansion and Regularized Zero-Forcing (RZF) are analyzed and compared in terms of Spectral Efficiency, and Achievable sum rate, a Rayleigh fading channel under perfect channel state information (CSI) is assumed. The template is used to format your paper and style the text. All margins, column widths, line spaces, and text fonts are prescribed; please do not alter them
Smartphone as an Agent of Anti-forensics: A Case of Workplace Environment in Kenya
Get PDFComputer anti-forensic techniques work to ensure that forensic evidence left behind after a digital crime is not easily uncovered by forensic investigators, if they are to uncover them, there will be a considerable delay. Smartphones have become a common device within an organization’s workforce where employees interact with highly confidential data that they access using their laptop computers at the workplace. This has led to the use of smartphones to commit digital crimes at the workplace. The primary objective of this study is to find out whether the use of smartphones at workplace environment in Kenya may be exploited to advance activities that may derail forensic investigations in the event of a digital crime. We also set to establish data security risks within organization and other techniques and/or methods by which smartphones may be used to exfiltrate data. Finally, we shall analyze research areas that require further attention from researchers to enhance defense and guard against smartphones data exfiltration. To achieve these objectives, we shall implement and test an android mobile software prototype, developed using android studio to send data exfiltration attempt to a web-based user interface when an employee within an organization uploads data above a set authorized limit. We shall review existing literature to understand other techniques that may be used to exfiltrate data from organizations as well as analyze research areas that require further attention from researchers to enhance defense and guard against data exfiltration through smartphones usage. We collected a total of two thousand five hundred and eighty-four records of data exfiltration attempts from our eleven sampled population. Of these records, One thousand eight hundred and ninety-one happened in the evening hours while six hundred and seven in the afternoon hours, then finally, eighty-six records were registered in the morning hours. In conclusion, the research study, has revealed that there exist challenges in reporting smartphone-based data exfiltration attempts while using the mobile-based software prototype.Data exfiltration attempts was observed to happen within organization’s workplace, with evening hours being the most affected by this vice with a figure of over one thousand data exfiltration attempts. We also noted that there exists, at least three categories of data security risks that organizations are exposed to when employees have their smartphones within the workplace. We recorded an additional eleven other techniques and methods by which a smartphone may be used to steal data from an organization
Predictive System for Heart Disease Using a Machine Learning Trained Model
Get PDFHeart as one of the essential organ of the human body and with its related disease such as cardiovascular diseases accounts for the death of many in our society over the last decades, and also regarded as one of the most life-threatening diseases in the world. Hence we seek to predict a system for Heart disease using a supervised Machine Learning (ML) trained model in MATLAB2018 workflow in a real-time environment. To develop the system, 299 heart sounds from patients were obtained and labeled as normal and abnormal heart sound. Features were extracted and labeled as dataset; K Nearest Neighbour (KNN), Support Vector Machine (SVM) and Decision Tree (DT) algorithm were used as the training platform. From the classification analysis developed using the supervised ML trained model in MATLAB2018 in conjunction with system software features for the prediction of the heartbeat for both current and predefined of a heart condition algorithms used in training the dataset for the prediction when principle component analysis was enabled, the result shows that KNN algorithm has the highest and best accuracy of 94.4%, followed by the SVM with 84.4% and DT had 81.1%. while from the evaluation analysis, KNN on Receive Operation Characteristic Curve (ROC) with 90% variance and training time of 12.88 seconds on positive class of abnormal over false classes of normal heart sound has AUC as 0.94 and on ROC curve with PCA 90% variance and training time of 1.7119 seconds on positive class of normal over negative classes of abnormal heart sound has AUC as 0.89 efficiency.
Hence the analysis from the result shows that out of the three classified algorithms used, KNN predicts and have the highest accuracy and is more efficient with respect to real-time environment
Predictive Human Resource Analytics Using Data Mining Classification Techniques
Get PDFThe turnover ratio of employees in organization is the most important concerns as employees switching of organization/ job leaves huge gape and affects the performance of that organization. Among many, job satisfaction is the prime reason of employees to quit/switch, which is also directly related to human resource management (HRM) practices of the organization. It is always difficult and sometime beyond the control of human resource (HR) department to retained their well-trained and skilled employees but Data mining can play role to predict those employees who are expected to quit/leave an organization such that the HR department can device intervention strategy or look for alternative. In this paper, we focus on similar problem, where we use data mining techniques such as J48, Naive Bayes, and Logistic Regression predict employees who will leave the organization. Our data consists of different indicator values and some other important features such as number of projects, supervisor evaluation score and experience. We show that J48 perform well with accuracy 98.84% and TP rate 0.984%. Conventional statistical analysis has been used in literature to identify important factors affecting employees satisfaction but there is not agreed set. We also apply data mining techniques to identify such factors using two approaches such as Bayesian Network and IR. Finally, we provide a decision tree based model for decision makers that can easily stimulate employees satisfaction level for better retention policy
Novel Resource Allocation Algorithm for TV White Space Networks Using Hybrid Firefly Algorithm
Get PDFThere is continued increased demand for dynamic spectrum access of TV White Spaces (TVWS) due to growing need for wireless broadband. Some of the use cases such as cellular (2G/3G/4G/5G) access to TVWS may have a high density of users that want to make use of TVWS. When there is a high of density secondary users (SUs) in a TVWS network, there is possibility of high interference among SUs that exceeds the desired threshold and also harmful interference to primary users (PUs). Optimization of resource allocation (power and spectrum allocation) is therefore necessary so as to protect the PUs against the harmful interference and to reduce the level of interference among SUs. In this paper, a novel and improved resource allocation algorithm based on hybrid firefly algorithm, genetic algorithm and particle swarm optimization (FAGAPSO) has been designed and applied for joint power and spectrum allocation. Computer simulations have been done using Matlab to validate the performance of the proposed algorithm. Simulation results show that compared to firefly algorithm (FA), particle swarm optimization (PSO) and genetic algorithm (GA), the algorithm improves the PU SINR, SU sum throughput and SU signal to interference noise (SINR) ratio in a TVWS network. Only one algorithm considered (SAP) has better PU SINR, SU sum throughput and SU signal to interference noise (SINR) ratio in a TVWS network but it has poor running time