Classification of Coronary Artery Disease Using Multilayer Perceptron Neural Network

Coronary artery disease (CAD) has been the leading cause of death worldwide over the past 10 years. Researchers have been using several data mining techniques to help healthcare professionals diagnose heart disease. The neural network (NN) can provide an excellent solution to identify and classify different diseases. The artificial neural network (ANN) methods play an essential role in recognizes diseases in the CAD. The authors proposed multilayer perceptron neural network (MLPNN) among one hidden layer neuron (MLP) and four hidden layers neurons (P-MLP)-based highly accurate artificial neural network (ANN) method for the classification of the CAD dataset. Therefore, the ten-fold cross-validation (T-FCV) method, P-MLP algorithms, and base classifiers of MLP were employed. The P-MLP algorithm yielded very high accuracy (86.47% in CAD-56 and 98.35% in CAD-59 datasets) and F1-Score (90.36% in CAD-56 and 98.83% in CAD-59 datasets) rates, which have not been reported simultaneously in the MLP.

Novel Discrete Rao Algorithms for Solving the Travelling Salesman Problem

Rao algorithms are population-based metaphor-less optimization algorithms. Rao algorithms consist of three algorithms characterized by three mathematical equations. These algorithms use the characteristics of the best and worst solution to modify the current population along with some characteristics of a random solution. These algorithms are found to be very efficient for continuous optimization problems. In this paper, efforts are made to convert Rao 1 algorithm to its discrete form. This paper proposes three techniques for converting these continuous Rao algorithms to their discrete form. One of the techniques is based on swap operator used for transforming PSO to discrete PSO, and the other two techniques are based on two novel mutating techniques. The algorithms are applied to symmetric TSP problems, and the results are compared with different state of the art algorithms, including discrete bat algorithm (DBA), discrete cuckoo search (DCS), ant colony algorithm, and GA. The results of Rao algorithms are highly competitive compared to the rest of the algorithms

A New Modeling Approach for the Video Pre-Analysis of Video Surveillance Systems

The work presented in this paper aims to develop a new architecture for video surveillance systems. Among the problems encountered when tracking and classifying objects are groups of occluded objects. Simplifying the representation of objects leads to other reliable object tracking with smaller amounts of information used but protection of the necessary characteristics. Therefore, modeling moving objects into a simpler form can be considered a pre-analysis technique. Objects can be represented in different ways, and the choice of the representation of an object strongly depends on the field of application. An example of a video surveillance system respecting this architecture and using the pre-analysis method is proposed.

Application of a Fuzzy MCDM Method to Select the Best Operating System for an Efficient Security-Aware Design of Embedded Systems

In this work, the authors present their solution to select the best operating system for an efficient security-aware design of embedded systems. This problem is formulated as a MCDM problem and solved using a hybrid approach combining fuzzy AHP and fuzzy VIKOR. This combination enables the authors to take profit of both methods. From AHP, they exploited the hierarchy and the pairwise comparison between criteria that leads to finding the importance (weight) of each criteria more consistently. On the other hand, from the VIKOR method, they leverage its power to compromise between conflictual criteria. Since they are dealing with unprecise and subjective advises, they opt for the fuzzy versions of the AHP and VIKOR methods dealing with triangular fuzzy numbers. They used fuzzy AHP to calculate weights of criteria which are served later as inputs for the fuzzy VIKOR method. The outcome of this work is to assist embedded designers to select the most appropriate embedded OS for efficient design of secure embedded systems.

Feature Selection and Knapsack Problem Resolution Based on a Discrete Backtracking Optimization Algorithm

Backtracking search optimization algorithm is a recent stochastic-based global search algorithm for solving real-valued numerical optimization problems. In this paper, a binary version of backtracking algorithm is proposed to deal with 0-1 optimization problems such as feature selection and knapsack problems. Feature selection is the process of selecting a subset of relevant features for use in model construction. Irrelevant features can negatively impact model performances. On the other hand, knapsack problem is a well-known optimization problem used to assess discrete algorithms. The objective of this research is to evaluate the discrete version of backtracking algorithm on the two mentioned problems and compare obtained results with other binary optimization algorithms using four usual classifiers: logistic regression, decision tree, random forest, and support vector machine. Empirical study on biological microarray data and experiments on 0-1 knapsack problems show the effectiveness of the binary algorithm and its ability to achieve good quality solutions for both problems.

COVID-19 Infection and Air Pollution Characteristics in USA

Coronavirus is diagnosed as a human-to-human infection at the initial stage by many of the researchers. As coronavirus is primarily targeting the respiratory system of the human body, the study tries to explore the relationship between pollution and increased number of cases in the states of the USA. The objective of the study is to determine whether the air quality index (AQI) of the year 2019 has a concern in the increasing number of coronavirus disease (COVID-19) cases in USA. This study included data of coronavirus confirmed and death cases from the dates January 22nd 2020 to June 30th 2020 for more than 45 states of the USA. Six AQI defining parameters—CO, NO2, Ozone (O3), PM10, PM2.5, and SO2—are selected for the study. The present study tried to identify whether air pollution is playing a significant role in spreading the coronavirus pandemic or not. Results confirmed that air quality index (AQI) defining parameters are not the sole factor behind the rise in the number of coronavirus cases in the USA.

Performance Estimation of Frequency Regulation for a Micro-Grid Power System Using PSO-PID Controller

This paper proposes the particle swarm optimization (PSO) technique-based proportional integral derivative (PID) controller suggested for frequency regulation of a micro grid (MG) system. MG system integrates with thermal power generating units, renewable energy sources (RES) like photovoltaic (PV), wind energy generators (WTG), and Energy storage systems (ESS) such as fuel cell (FC) and battery energy storage system (BESS). Indentifying the supremacy of proposed technique-based controller and supremacy is examined with three objective functions (integral absolute error [IAE], integral time absolute error [ITAE], and integral squared error [ISE]). The results of the system are compared with conventional PID controller results. From the comparison, it is clearly evident that PSO-PID controller gives better performance over conventional methods in terms of various time domain specific parameters such as settling time, peak overshoot, and undershoot. In both methods, ITAE objective function used controller produce more effective response in MG under sudden load demand situation.

Asymmetric Encryption Scheme to Protect Cloud Data Using Paillier-Cryptosystem

Cloud computing is used for large shared resources to facilitate execution and storage. So there is a need of resolving crucial security issues to avoid data theft. Hence cloud security provides data encryption for security parameters to change plain-text to cipher-text. The homomorphic encryption technique is used for performing operations on encrypted data. To manage the huge and growing informational collections that are being prepared these days, great encryption execution is a significant advance for the common sense of homomorphic encryption techniques, the Paillier cryptosystem is also used by researchers for securing the decimal digits of information. In the present work, a hybrid Paillier cryptosystem technique is used for reducing the bit length of the cipher-text by performing hex code operations on encryption. The proposed method has been implemented in the use of two object-oriented programming languages i.e. C++ and Python programming languages. The simulated results show the minimum encrypted bit length as well as provide more secure data. And we have also analyzed our algorithm based on the two parameters namely space complexity and time complexity which are represented in the form of tables and graphs given below.

Effects of Class Imbalance Using Machine Learning Algorithms

Class imbalance is the major hurdle for machine learning-based systems. Data set is the backbone of machine learning and must be studied to handle the class imbalance. The purpose of this paper is to investigate the effect of class imbalance on the data sets. The proposed methodology determines the model accuracy for class distribution. To find possible solutions, the behaviour of an imbalanced data set was investigated. The study considers two case studies with data set divided balanced to unbalanced class distribution. Testing of the data set with trained and test data was carried out for standard machine learning algorithms. Model accuracy for class distribution was measured with the training data set. Further, the built model was tested with individual binary class. Results show that, for the improvement of the system performance, it is essential to work on class imbalance problems. The study concludes that the system produces biased results due to the majority class. In the future, the multiclass imbalance problem can be studied using advanced algorithms.

Identification and Evaluation of Homographic Puns Using Similarity Methods

Natural language processing is a subfield of linguistics concerned with the interactions between computers and human language, specifically in how to program computers to process and analyze large amounts of text data (natural language data). WSD, word sense disambiguation in natural language processing, is the task of determining the correct annotation of the pun word in given context. This paper describes about the endeavor in using cosine similarity method for detection of a single homographic pun in given context, its location, and the correct annotation with respect to helping words in the context. This paper includes two approaches: BIT_SYS1 and BIT_SYS2. The first contains the words having synset count one as it cannot be pun but it can serve as helping word to the pun, and in the later words with synset count one is eliminated and the concept of helping word is abandoned. Performance of BIT_SYS2 is better than BIT_SYS1 as F1 score of BIT_SYS2(0.8571, 1.0000, 1.0000) is higher than BIT_SYS1(0.8439, 0.8648, 0.8648) in pun detection task, pun location task, and pun annotation task.