Abstract: In the field of machine learning, the ensemble has been employed as a common methodology to improve the performance upon multiple base classifiers. However, the true predictions are often canceled out by the false ones during consensus due to a phenomenon called “curse of correlation” which is represented as the strong interferences among the predictions produced by the base classifiers. In addition, the existing practices are still not able to effectively mitigate the problem of imbalanced classification. Based on the analysis on our experiment results, we conclude that the two problems are caused by some inherent deficiencies in the approach of consensus. Therefore, we create an enhanced ensemble algorithm which adopts a designed rank-based chain-mode consensus to overcome the two problems. In order to evaluate the proposed ensemble algorithm, we employ a well-known benchmark data set NSL-KDD (the improved version of dataset KDDCup99 produced by University of New Brunswick) to make comparisons between the proposed and 8 common ensemble algorithms. Particularly, each compared ensemble classifier uses the same 22 base classifiers, so that the differences in terms of the improvements toward the accuracy and reliability upon the base classifiers can be truly revealed. As a result, the proposed rank-based chain-mode consensus is proved to be a more effective ensemble solution than the traditional consensus approach, which outperforms the 8 ensemble algorithms by 20% on almost all compared metrices which include accuracy, precision, recall, F1-score and area under receiver operating characteristic curve.
Abstract: Vehicular Adhoc Networks (VANETs), a subset of Mobile Adhoc Networks (MANETs), refers to a set of smart vehicles used for road safety. This vehicle provides communication services among one another or with the Road Side Unit (RSU). Security is one of the most critical issues related to VANET as the information transmitted is distributed in an open access environment. As each vehicle is not a source of all messages, most of the communication depends on the information received from other vehicles. To protect VANET from malicious action, each vehicle must be able to evaluate, decide and react locally on the information received from other vehicles. Therefore, message verification is more challenging in VANET because of the security and privacy concerns of the participating vehicles. To overcome security threats, we propose Monitoring Algorithm that detects malicious nodes based on the pre-selected threshold value. The threshold value is compared with the distrust value which is inherently tagged with each vehicle. The proposed Monitoring Algorithm not only detects malicious vehicles, but also isolates the malicious vehicles from the network. The proposed technique is simulated using Network Simulator2 (NS2) tool. The simulation result illustrated that the proposed Monitoring Algorithm outperforms the existing algorithms in terms of malicious node detection, network delay, packet delivery ratio and throughput, thereby uplifting the overall performance of the network.
Abstract: This paper employs a heuristic algorithm to solve engineering problems including truss structure optimization and optimal chiller loading (OCL) problems. Two different type algorithms, real-valued differential evolution (DE) and modified binary differential evolution (MBDE), are successfully integrated and then can obtain better performance in solving engineering problems. In order to demonstrate the performance of the proposed algorithm, this study adopts each one testing case of truss structure optimization and OCL problems to compare the results of other heuristic optimization methods. The result indicates that the proposed algorithm can obtain similar or better solution in comparing with previous studies.
Abstract: The smart antenna is the prominent technology that has become known in recent years to meet the growing demands of wireless communications. In an overcrowded atmosphere, its application is growing gradually. A methodical evaluation of the performance of Fixed Beamforming algorithms for smart antennas such as Multiple Sidelobe Canceller (MSC), Maximum Signal-to-interference ratio (MSIR) and minimum variance (MVDR) has been comprehensively presented in this paper. Simulation results show that beamforming is helpful in providing optimized response towards desired directions. MVDR beamformer provides the most optimal solution.
Abstract: An Upgraded Cuckoo Search Algorithm is proposed here to solve optimization problems based on the improvements made in the earlier versions of Cuckoo Search Algorithm. Short comings of the earlier versions like slow convergence, trap in local optima improved in the proposed version by random initialization of solution by suggesting an Improved Lambda Iteration Relaxation method, Random Gaussian Distribution Walk to improve local search and further proposing Greedy Selection to accelerate to optimized solution quickly and by “Study Nearby Strategy” to improve global search performance by avoiding trapping to local optima. It is further proposed to generate better solution by Crossover Operation. The proposed strategy used in algorithm shows superiority in terms of high convergence speed over several classical algorithms. Three standard algorithms were tested on a 6-generator standard test system and the results are presented which clearly demonstrate its superiority over other established algorithms. The algorithm is also capable of handling higher unit systems.
Abstract: Learning from very big datasets is a significant problem for most present data mining and machine learning algorithms. MicroRNA (miRNA) is one of the important big genomic and non-coding datasets presenting the genome sequences. In this paper, a hybrid method for the classification of the miRNA data is proposed. Due to the variety of cancers and high number of genes, analyzing the miRNA dataset has been a challenging problem for researchers. The number of features corresponding to the number of samples is high and the data suffer from being imbalanced. The feature selection method has been used to select features having more ability to distinguish classes and eliminating obscures features. Afterward, a Convolutional Neural Network (CNN) classifier for classification of cancer types is utilized, which employs a Genetic Algorithm to highlight optimized hyper-parameters of CNN. In order to make the process of classification by CNN faster, Graphics Processing Unit (GPU) is recommended for calculating the mathematic equation in a parallel way. The proposed method is tested on a real-world dataset with 8,129 patients, 29 different types of tumors, and 1,046 miRNA biomarkers, taken from The Cancer Genome Atlas (TCGA) database.
Abstract: High Voltage Substations (HVS) are the intermediate step between production of power and successfully transmitting it to clients, making them one of the most important checkpoints in power grids. Nowadays - renewable resources and consequently distributed generation are growing fast, the construction of HVS is of high importance both in terms of quality and time completion so that new energy producers can quickly and safely intergrade in power grids. The resources needed, such as machines and workers, should be carefully allocated so that the construction of a HVS is completed on time, with the lowest possible cost (e.g. not spending additional cost that were not taken into consideration, because of project delays), but in the highest quality. In addition, there are milestones and several checkpoints to be precisely achieved during construction to ensure the cost and timeline control and to ensure that the percentage of governmental funding will be granted. The management of such a demanding project is a NP-hard problem that consists of prerequisite constraints and resource limits for each task of the project. In this work, a hybrid meta-heuristic method is implemented to solve this problem. Meta-heuristics have been proven to be quite useful when dealing with high-dimensional constraint optimization problems. Hybridization of them results in boost of their performance.
Abstract: E-learning platforms, such as Blackboard have two major shortcomings: limited data capture as a result of the limitations of SCORM (Shareable Content Object Reference Model), and lack of incorporation of Artificial Intelligence (AI) and machine learning algorithms which could lead to better course adaptations. With the recent development of Experience Application Programming Interface (xAPI), a large amount of additional types of data can be captured and that opens a window of possibilities from which online education can benefit. In a corporate setting, where companies invest billions on the learning and development of their employees, some learner behaviours can be troublesome for they can hinder the knowledge development of a learner. Behaviours that hinder the knowledge development also raise ambiguity about learner’s knowledge mastery, specifically those related to gaming the system. Furthermore, a company receives little benefit from their investment if employees are passing courses without possessing the required knowledge and potential compliance risks may arise. Using xAPI and rules derived from a state-of-the-art review, we identified three learner behaviours, primarily related to guessing, in a corporate compliance course. The identified behaviours are: trying each option for a question, specifically for multiple-choice questions; selecting a single option for all the questions on the test; and continuously repeating tests upon failing as opposed to going over the learning material. These behaviours were detected on learners who repeated the test at least 4 times before passing the course. These findings suggest that gauging the mastery of a learner from multiple-choice questions test scores alone is a naive approach. Thus, next steps will consider the incorporation of additional data points, knowledge estimation models to model knowledge mastery of a learner more accurately, and analysis of the data for correlations between knowledge development and identified learner behaviours. Additional work could explore how learner behaviours could be utilised to make changes to a course. For example, course content may require modifications (certain sections of learning material may be shown to not be helpful to many learners to master the learning outcomes aimed at) or course design (such as the type and duration of feedback).
Abstract: This paper presents a Machine Learning (ML) approach to support Meningitis diagnosis in patients at a children’s hospital in Sao Paulo, Brazil. The aim is to use ML techniques to reduce the use of invasive procedures, such as cerebrospinal fluid (CSF) collection, as much as possible. In this study, we focus on predicting the probability of Meningitis given the results of a blood and urine laboratory tests, together with the analysis of pain or other complaints from the patient. We tested a number of different ML algorithms, including: Adaptative Boosting (AdaBoost), Decision Tree, Gradient Boosting, K-Nearest Neighbors (KNN), Logistic Regression, Random Forest and Support Vector Machines (SVM). Decision Tree algorithm performed best, with 94.56% and 96.18% accuracy for training and testing data, respectively. These results represent a significant aid to doctors in diagnosing Meningitis as early as possible and in preventing expensive and painful procedures on some children.
Abstract: Arabic offline handwriting recognition systems are considered as one of the most challenging topics. Arabic Handwritten Numeral Strings are used to automate systems that deal with numbers such as postal code, banking account numbers and numbers on car plates. Segmentation of connected numerals is the main bottleneck in the handwritten numeral recognition system. This is in turn can increase the speed and efficiency of the recognition system. In this paper, we proposed algorithms for automatic segmentation and feature extraction of Arabic handwritten numeral strings based on Watershed approach. The algorithms have been designed and implemented to achieve the main goal of segmenting and extracting the string of numeral digits written by hand especially in a courtesy amount of bank checks. The segmentation algorithm partitions the string into multiple regions that can be associated with the properties of one or more criteria. The numeral extraction algorithm extracts the numeral string digits into separated individual digit. Both algorithms for segmentation and feature extraction have been tested successfully and efficiently for all types of numerals.
Abstract: Nature-inspired metaheuristic algorithms, particularly those founded on swarm intelligence, have attracted much attention over the past decade. Firefly algorithm has appeared in approximately seven years ago, its literature has enlarged considerably with different applications. It is inspired by the behavior of fireflies. The aim of this paper is the application of firefly algorithm for solving a nonlinear algebraic system. This resolution is needed to study the Selective Harmonic Eliminated Pulse Width Modulation strategy (SHEPWM) to eliminate the low order harmonics; results have been applied on multilevel inverters. The final results from simulations indicate the elimination of the low order harmonics as desired. Finally, experimental results are presented to confirm the simulation results and validate the efficaciousness of the proposed approach.
Abstract: In the first decades of the 21st century, in the electronic trading environment, algorithmic capital investments became the primary tool to make a profit by speculations in financial markets. A significant number of traders, private or institutional investors are participating in the capital markets every day using automated algorithms. The autonomous trading software is today a considerable part in the business intelligence system of any modern financial activity. The trading decisions and orders are made automatically by computers using different mathematical models. This paper will present one of these models called Price Prediction Line. A mathematical algorithm will be revealed to build a reliable trend line, which is the base for limit conditions and automated investment signals, the core for a computerized investment system. The paper will guide how to apply these tools to generate entry and exit investment signals, limit conditions to build a mathematical filter for the investment opportunities, and the methodology to integrate all of these in automated investment software. The paper will also present trading results obtained for the leading German financial market index with the presented methods to analyze and to compare different automated investment algorithms. It was found that a specific mathematical algorithm can be optimized and integrated into an automated trading system with good and sustained results for the leading German Market. Investment results will be compared in order to qualify the presented model. In conclusion, a 1:6.12 risk was obtained to reward ratio applying the trigonometric method to the DAX Deutscher Aktienindex on 24 months investment. These results are superior to those obtained with other similar models as this paper reveal. The general idea sustained by this paper is that the Price Prediction Line model presented is a reliable capital investment methodology that can be successfully applied to build an automated investment system with excellent results.
Abstract: This paper aims to demonstrate how various algorithms can be implemented within swarms of autonomous robots to provide continuous inspection within underground pipeline networks. Current methods of fault detection within pipes are costly, time consuming and inefficient. As such, solutions tend toward a more reactive approach, repairing faults, as opposed to proactively seeking leaks and blockages. The paper presents an efficient inspection method, showing that autonomous swarm robotics is a viable way of monitoring underground infrastructure. Tailored adaptations of various Vehicle Routing Problems (VRP) and path-planning algorithms provide a customised inspection procedure for complicated networks of underground pipes. The performance of multiple algorithms is compared to determine their effectiveness and feasibility. Notable inspirations come from ant colonies and stigmergy, graph theory, the k-Chinese Postman Problem ( -CPP) and traffic theory. Unlike most swarm behaviours which rely on fast communication between agents, underground pipe networks are a highly challenging communication environment with extremely limited communication ranges. This is due to the extreme variability in the pipe conditions and relatively high attenuation of acoustic and radio waves with which robots would usually communicate. This paper illustrates how to optimise the inspection process and how to increase the frequency with which the robots pass each other, without compromising the routes they are able to take to cover the whole network.
Abstract: Since its inception, predictive analysis has revolutionized the IT industry through its robustness and decision-making facilities. It involves the application of a set of data processing techniques and algorithms in order to create predictive models. Its principle is based on finding relationships between explanatory variables and the predicted variables. Past occurrences are exploited to predict and to derive the unknown outcome. With the advent of big data, many studies have suggested the use of predictive analytics in order to process and analyze big data. Nevertheless, they have been curbed by the limits of classical methods of predictive analysis in case of a large amount of data. In fact, because of their volumes, their nature (semi or unstructured) and their variety, it is impossible to analyze efficiently big data via classical methods of predictive analysis. The authors attribute this weakness to the fact that predictive analysis algorithms do not allow the parallelization and distribution of calculation. In this paper, we propose to extend the predictive analysis algorithm, Classification And Regression Trees (CART), in order to adapt it for big data analysis. The major changes of this algorithm are presented and then a version of the extended algorithm is defined in order to make it applicable for a huge quantity of data.
Abstract: Geometric modeling plays an important role in the
constructions and manufacturing of curve, surface and solid
modeling. Their algorithms are critically important not only in
the automobile, ship and aircraft manufacturing business, but are
also absolutely necessary in a wide variety of modern applications,
e.g., robotics, optimization, computer vision, data analytics and
visualization. The calculation and display of geometric objects
can be accomplished by these six techniques: Polynomial basis,
Recursive, Iterative, Coefficient matrix, Polar form approach and
Pyramidal algorithms. In this research, the coefficient matrix (simply
called monomial form approach) will be used to model polynomial
rectangular patches, i.e., Said-Ball, Wang-Ball, DP, Dejdumrong and
NB1 surfaces. Some examples of the monomial forms for these
surface modeling are illustrated in many aspects, e.g., construction,
derivatives, model transformation, degree elevation and degress
reduction.
Abstract: Newton-Lagrange Interpolations are widely used in
numerical analysis. However, it requires a quadratic computational
time for their constructions. In computer aided geometric design
(CAGD), there are some polynomial curves: Wang-Ball, DP and
Dejdumrong curves, which have linear time complexity algorithms.
Thus, the computational time for Newton-Lagrange Interpolations
can be reduced by applying the algorithms of Wang-Ball, DP and
Dejdumrong curves. In order to use Wang-Ball, DP and Dejdumrong
algorithms, first, it is necessary to convert Newton-Lagrange
polynomials into Wang-Ball, DP or Dejdumrong polynomials. In
this work, the algorithms for converting from both uniform and
non-uniform Newton-Lagrange polynomials into Wang-Ball, DP and
Dejdumrong polynomials are investigated. Thus, the computational
time for representing Newton-Lagrange polynomials can be reduced
into linear complexity. In addition, the other utilizations of using
CAGD curves to modify the Newton-Lagrange curves can be taken.
Abstract: 2016 has become the year of the Artificial Intelligence explosion. AI technologies are getting more and more matured that most world well-known tech giants are making large investment to increase the capabilities in AI. Machine learning is the science of getting computers to act without being explicitly programmed, and deep learning is a subset of machine learning that uses deep neural network to train a machine to learn features directly from data. Deep learning realizes many machine learning applications which expand the field of AI. At the present time, deep learning frameworks have been widely deployed on servers for deep learning applications in both academia and industry. In training deep neural networks, there are many standard processes or algorithms, but the performance of different frameworks might be different. In this paper we evaluate the running performance of two state-of-the-art distributed deep learning frameworks that are running training calculation in parallel over multi GPU and multi nodes in our cloud environment. We evaluate the training performance of the frameworks with ResNet-50 convolutional neural network, and we analyze what factors that result in the performance among both distributed frameworks as well. Through the experimental analysis, we identify the overheads which could be further optimized. The main contribution is that the evaluation results provide further optimization directions in both performance tuning and algorithmic design.
Abstract: Seizure is the main factor that affects the quality of life of epileptic patients. The diagnosis of epilepsy, and hence the identification of epileptogenic zone, is commonly made by using continuous Electroencephalogram (EEG) signal monitoring. Seizure identification on EEG signals is made manually by epileptologists and this process is usually very long and error prone. The aim of this paper is to describe an automated method able to detect seizures in EEG signals, using knowledge discovery in database process and data mining methods and algorithms, which can support physicians during the seizure detection process. Our detection method is based on Artificial Neural Network classifier, trained by applying the multilayer perceptron algorithm, and by using a software application, called Training Builder that has been developed for the massive extraction of features from EEG signals. This tool is able to cover all the data preparation steps ranging from signal processing to data analysis techniques, including the sliding window paradigm, the dimensionality reduction algorithms, information theory, and feature selection measures. The final model shows excellent performances, reaching an accuracy of over 99% during tests on data of a single patient retrieved from a publicly available EEG dataset.
Abstract: In general, issues related to design and maintenance are considered in an independent manner. However, the decisions made in these two sets influence each other. The design for maintenance is considered an opportunity to optimize the life cycle cost of a product, particularly in the nuclear or aeronautical field, where maintenance expenses represent more than 60% of life cycle costs. The design of large-scale systems starts with product architecture, a choice of components in terms of cost, reliability, weight and other attributes, corresponding to the specifications. On the other hand, the design must take into account maintenance by improving, in particular, real-time monitoring of equipment through the integration of new technologies such as connected sensors and intelligent actuators. We noticed that different approaches used in the Design For Maintenance (DFM) methods are limited to the simultaneous characterization of the reliability and maintainability of a multi-component system. This article proposes a method of DFM that assists designers to propose dynamic maintenance for multi-component industrial systems. The term "dynamic" refers to the ability to integrate available monitoring data to adapt the maintenance decision in real time. The goal is to maximize the availability of the system at a given life cycle cost. This paper presents an approach for simultaneous optimization of the design and maintenance of multi-component systems. Here the design is characterized by four decision variables for each component (reliability level, maintainability level, redundancy level, and level of monitoring data). The maintenance is characterized by two decision variables (the dates of the maintenance stops and the maintenance operations to be performed on the system during these stops). The DFM model helps the designers choose technical solutions for the large-scale industrial products. Large-scale refers to the complex multi-component industrial systems and long life-cycle, such as trains, aircraft, etc. The method is based on a two-level hybrid algorithm for simultaneous optimization of design and maintenance, using genetic algorithms. The first level is to select a design solution for a given system that considers the life cycle cost and the reliability. The second level consists of determining a dynamic and optimal maintenance plan to be deployed for a design solution. This level is based on the Maintenance Free Operating Period (MFOP) concept, which takes into account the decision criteria such as, total reliability, maintenance cost and maintenance time. Depending on the life cycle duration, the desired availability, and the desired business model (sales or rental), this tool provides visibility of overall costs and optimal product architecture.
Abstract: Acoustic noise becoming ever more obnoxious radiated by voltage source inverter fed induction motor drive in modern and industrial applications. The drive utilized for industrial and modern applications should use “spread spectrum” innovation known as Random pulse width modulation (PWM) algorithms where acoustic noise emanates through the machine should be critically concerned. This paper illustrates three types of random PWM control algorithms with fixed switching frequency namely 1) Random modulating PWM 2) Random carrier PWM and 3) Random modulating-carrier PWM. The spectrum plots of the motor stator current demonstrate the strength and robustness of the proposed PWM algorithms. To affirm the proposed algorithms, experimental tests have been conducted using dSPACE rt1104 control board on a v/f control three phase induction motor drive fed by DC link cascaded multilevel inverter.