Search results for: machine-learning algorithms

Authors: Daham Owaid Matrood, Naqaa Hussein Raheem

Abstract:

Evolutionary algorithms and Artificial neural networks (ANN) are two relatively young research areas that were subject to a steadily growing interest during the past years. This paper examines the use of Particle Swarm Optimization (PSO) to train a multi-layer feed forward neural network for demand forecasting. We use in this paper weekly demand data for packed cement and towels, which have been outfitted by the Northern General Company for Cement and General Company of prepared clothes respectively. The results showed superiority of trained neural networks using particle swarm optimization on neural networks trained using error back propagation because their ability to escape from local optima.

Keywords: artificial neural network, demand forecasting, particle swarm optimization, weight optimization

Procedia PDF Downloads 452

Authors: Egor E. Nuzhin, Nikolay V. Brilliantov

Abstract:

We propose an approach for addressing Reinforcement Learning (RL) problems. It combines the ideas of umbrella sampling, borrowed from Monte Carlo technique of computational physics and chemistry, with optimal control methods, and is realized on the base of neural networks. This results in a powerful algorithm, designed to solve hard RL problems – the problems, with long-time delayed reward, state-traps sticking and a lack of terminal states. It outperforms the prominent algorithms, such as PPO, RND, iLQR and VI, which are among the most efficient for the hard problems. The new algorithm deals with a continuous ensemble of agents and expected return, that includes the ensemble entropy. This results in a quick and efficient search of the optimal policy in terms of ”exploration-exploitation trade-off” in the state-action space.

Keywords: umbrella sampling, reinforcement learning, policy gradient, dynamic programming

Procedia PDF Downloads 21

Authors: Omar Dib, Alexandre Caminada, Marie-Ange Manier

Abstract:

The purpose of this study is to introduce a novel approach to solve the one-to-one shortest path problem. A directed connected graph is assumed in which all edges’ weights are positive. Our method is based on a memetic algorithm in which we combine a genetic algorithm (GA) and a variable neighborhood search method (VNS). We compare our approximate method with two exact algorithms Dijkstra and Integer Programming (IP). We made experimentations using random generated, complete and real graph instances. In most case studies, numerical results show that our method outperforms exact methods with 5% average gap to the optimality. Our algorithm’s average speed is 20-times faster than Dijkstra and more than 1000-times compared to IP. The details of the experimental results are also discussed and presented in the paper.

Keywords: shortest path problem, Dijkstra’s algorithm, integer programming, memetic algorithm

Procedia PDF Downloads 467

Authors: Sagar Hedaoo, Fazle Baki, Ahmed Azab

Abstract:

In supply chain management, network design for scalable manufacturing facilities is an emerging field of research. Facility location allocation assigns facilities to customers to optimize the overall cost of the supply chain. To further optimize the costs, capacities of these facilities can be changed in accordance with customer demands. A mathematical model is formulated to fully express the problem at hand and to solve small-to-mid range instances. A dedicated constraint has been developed to restrict emissions in line with the Kyoto protocol. This problem is NP-Hard; hence, a simulated annealing metaheuristic has been developed to solve larger instances. A case study on the USA-Canada cross border crossing is used.

Keywords: emission, mixed integer linear programming, metaheuristic, simulated annealing

Procedia PDF Downloads 309

Authors: Rashi Jain

Abstract:

The LTE systems employ Orthogonal Frequency Division Multiplexing (OFDM) as the multiple access technology for the Downlink channels. For enhanced performance, accurate channel estimation is required. Various algorithms such as Least Squares (LS), Minimum Mean Square Error (MMSE) and Recursive Least Squares (RLS) can be employed for the purpose. The paper proposes channel estimation algorithm based on Kalman Filter for LTE-Downlink system. Using the frequency domain pilots, the initial channel response is obtained using the LS criterion. Then Kalman Filter is employed to track the channel variations in time-domain. To suppress the noise within a symbol, threshold processing is employed. The paper draws comparison between the LS, MMSE, RLS and Kalman filter for channel estimation. The parameters for evaluation are Bit Error Rate (BER), Mean Square Error (MSE) and run-time.

Keywords: LTE, channel estimation, OFDM, RLS, Kalman filter, threshold

Procedia PDF Downloads 356

Authors: Zhongfei Cai, Hui He, Changsheng Li

Abstract:

This paper studied in the automatic repeat request (ARQ) transmission technique in Mars detection telecommunications system. An ARQ method applied to proximity-1 space link protocol was proposed by this paper. In order to ensure the efficiency of data reliable transmission, this ARQ method combined these different ARQ maneuvers characteristics. Considering the Mars detection communication environments, this paper analyzed the characteristics of the saturation throughput rate, packet dropping probability, average delay and energy efficiency with different ARQ algorithms. Combined thus results with the theories of ARQ transmission technique, an ARQ transmission project in Mars detection telecommunications system was established. The simulation results showed that this algorithm had excellent saturation throughput rate and energy efficiency with low complexity.

Keywords: ARQ, mars, CCSDS, proximity-1, deepspace

Procedia PDF Downloads 340

Authors: El Asnaoui Khalid, Chawki Youness, Aksasse Brahim, Ouanan Mohammed

Abstract:

In these recent decades, the important and fast growth in the development and demand of multimedia products is contributing to an insufficient in the bandwidth of device and network storage memory. Consequently, the theory of data compression becomes more significant for reducing the data redundancy in order to save more transfer and storage of data. In this context, this paper addresses the problem of the lossless and the near-lossless compression of images. This proposed method is based on Block SVD Power Method that overcomes the disadvantages of Matlab's SVD function. The experimental results show that the proposed algorithm has a better compression performance compared with the existing compression algorithms that use the Matlab's SVD function. In addition, the proposed approach is simple and can provide different degrees of error resilience, which gives, in a short execution time, a better image compression.

Keywords: image compression, SVD, block SVD power method, lossless compression, near lossless

Procedia PDF Downloads 387

Authors: Najam Aziz, Nasru Minallah, Ahmad Junaid, Kashaf Gul

Abstract:

Landcover classification using automated classification techniques, while employing remotely sensed multi-spectral imagery, is one of the promising areas of research. Different land conditions at different time are captured through satellite and monitored by applying different classification algorithms in specific environment. In this paper, a SPOT-5 image provided by SUPARCO has been studied and classified in Environment for Visual Interpretation (ENVI), a tool widely used in remote sensing. Then, Artificial Neural Network (ANN) classification technique is used to detect the land cover changes in Abbottabad district. Obtained results are compared with a pixel based Distance classifier. The results show that ANN gives the better overall accuracy of 99.20% and Kappa coefficient value of 0.98 over the Mahalanobis Distance Classifier.

Keywords: landcover classification, artificial neural network, remote sensing, SPOT 5

Procedia PDF Downloads 546

Authors: A. G. Kovalenko, Y. N. Amirgaliyev, A. U. Kalizhanova, L. S. Balgabayeva, A. H. Kozbakova, Z. S. Aitkulov

Abstract:

It is known that statistical information on operation of the compound multisite system is often far from the description of actual state of the system and does not allow drawing any conclusions about the correctness of its operation. For example, from the world practice of operation of systems of water supply, water disposal, it is known that total measurements at consumers and at suppliers differ between 40-60%. It is connected with mathematical measure of inaccuracy as well as ineffective running of corresponding systems. Analysis of widely-distributed systems is more difficult, in which subjects, which are self-maintained in decision-making, carry out economic interaction in production, act of purchase and sale, resale and consumption. This work analyzed mathematical models of sellers, consumers, arbitragers and the models of their interaction in the provision of dispersed single-product market of perfect competition. On the basis of these models, the methods, allowing estimation of every subject’s operating options and systems as a whole are given.

Keywords: dispersed systems, models, hydraulic network, algorithms

Procedia PDF Downloads 284

Authors: Rodrigo Aguiar, Adelino Ferreira

Abstract:

Road traffic accidents are the leading cause of unnatural death and injuries worldwide, representing a significant problem of road safety. In this context, the use of artificial intelligence with advanced machine learning techniques has gained prominence as a promising approach to predict traffic accidents. This article investigates the application of machine learning algorithms to develop traffic accident frequency prediction models. Models are evaluated based on performance metrics, making it possible to do a comparative analysis with traditional prediction approaches. The results suggest that machine learning can provide a powerful tool for accident prediction, which will contribute to making more informed decisions regarding road safety.

Keywords: machine learning, artificial intelligence, frequency of accidents, road safety

Procedia PDF Downloads 89

Authors: Raana Babadi Fathipour

Abstract:

Strict prerequisites of logical diaries united ought to demonstrate the exploratory information is (in)significant from the statistical point of view and has driven a soak increment within the utilization and advancement of the factual program. It is essential that the utilization of numerical and measurable strategies, counting chemometrics and many other factual methods/algorithms in nourishment science and innovation has expanded steeply within the final 20 a long time. Computational apparatuses accessible can be utilized not as it were to run factual investigations such as univariate and bivariate tests as well as multivariate calibration and improvement of complex models but also to run reenactments of distinctive scenarios considering a set of inputs or essentially making expectations for particular information sets or conditions. Conducting a fast look within the most legitimate logical databases (Pubmed, ScienceDirect, Scopus), it is conceivable to watch that measurable strategies have picked up a colossal space in numerous regions.

Keywords: food science, food technology, food safety, computational tools

Procedia PDF Downloads 67

Authors: M. Z. Kurian, M. V. Chidananda Murthy, H. S. Guruprasad

Abstract:

An approach to super-resolve the low-resolution (LR) image is presented in this paper which is very useful in multimedia communication, medical image enhancement and satellite image enhancement to have a clear view of the information in the image. The proposed Advanced B-Spline method generates a high-resolution (HR) image from single LR image and tries to retain the higher frequency components such as edges in the image. This method uses B-Spline technique and Crispening. This work is evaluated qualitatively and quantitatively using Mean Square Error (MSE) and Peak Signal to Noise Ratio (PSNR). The method is also suitable for real-time applications. Different combinations of decimation and super-resolution algorithms in the presence of different noise and noise factors are tested.

Keywords: advanced b-spline, image super-resolution, mean square error (MSE), peak signal to noise ratio (PSNR), resolution down converter

Procedia PDF Downloads 399

Authors: Endrick Barnacin, Jean-Luc Henry, Jimmy Nagau, Jack Molinie

Abstract:

Palynology is a field of interest in many disciplines due to its multiple applications: chronological dating, climatology, allergy treatment, and honey characterization. Unfortunately, the analysis of a pollen slide is a complicated and time consuming task that requires the intervention of experts in the field, which are becoming increasingly rare due to economic and social conditions. That is why the need for automation of this task is urgent. A lot of studies have investigated the subject using different standard image processing descriptors and sometimes hand-crafted ones.In this work, we make a comparative study between classical feature extraction methods (Shape, GLCM, LBP, and others) and Deep Learning (CNN, Autoencoders, Transfer Learning) to perform a recognition task over 80 regional pollen species. It has been found that the use of Transfer Learning seems to be more precise than the other approaches

Keywords: pollens identification, features extraction, pollens classification, automated palynology

Procedia PDF Downloads 136

Authors: Fares Hedayatit, Arnauld Joly, Panagiotis Papadimitriou

Abstract:

Many machine learning tasks such as text annotation usually require training over very big datasets, e.g., millions of web documents, that can be represented in a sparse input space. State-of the-art tree-based ensemble algorithms cannot scale to such datasets, since they include operations whose running time is a function of the input space size rather than a function of the non-zero input elements. In this paper, we propose an efficient splitting algorithm to leverage input sparsity within decision tree methods. Our algorithm improves training time over sparse datasets by more than two orders of magnitude and it has been incorporated in the current version of scikit-learn.org, the most popular open source Python machine learning library.

Keywords: big data, sparsely representable data, tree-based models, scalable learning

Procedia PDF Downloads 263

Authors: Wahiba Mohguen, Raïs El’hadi Bekka

Abstract:

This paper presents a denoising method called EMD-Custom that was based on Empirical Mode Decomposition (EMD) and the modified Customized Thresholding Function (Custom) algorithms. EMD was applied to decompose adaptively a noisy signal into intrinsic mode functions (IMFs). Then, all the noisy IMFs got threshold by applying the presented thresholding function to suppress noise and to improve the signal to noise ratio (SNR). The method was tested on simulated data and real ECG signal, and the results were compared to the EMD-Based signal denoising methods using the soft and hard thresholding. The results showed the superior performance of the proposed EMD-Custom denoising over the traditional approach. The performances were evaluated in terms of SNR in dB, and Mean Square Error (MSE).

Keywords: customized thresholding, ECG signal, EMD, hard thresholding, soft-thresholding

Procedia PDF Downloads 302

Authors: Mohammad Reza Akhavan Khaleghi

Abstract:

This paper shows changes done to the Reduced Finite Element Method (RFEM) that its result will be the most powerful numerical method that has been proposed so far (some forms of this method are so powerful that they can approximate the most complex equations simply Laplace equation!). Finite Element Method (FEM) is a powerful numerical method that has been used successfully for the solution of the existing problems in various scientific and engineering fields such as its application in CFD. Many algorithms have been expressed based on FEM, but none have been used in popular CFD software. In this section, full monopoly is according to Finite Volume Method (FVM) due to better efficiency and adaptability with the physics of problems in comparison with FEM. It doesn't seem that FEM could compete with FVM unless it was fundamentally changed. This paper shows those changes and its result will be a powerful method that has much better performance in all subjects in comparison with FVM and another computational method. This method is not to compete with the finite volume method but to replace it.

Keywords: reduced finite element method, new computational package, new finite element formulation, new higher-order form, new isogeometric analysis

Procedia PDF Downloads 118

Authors: Donald Wright

Abstract:

The ultimate goal of this study is to reinvigorate the philosophical underpinnings the study of urbanization with scientific data with the goal of circumventing what seems an inevitable future clash between rural and urban populations. To that end urban infrastructure must be sustainable economically, politically and ecologically over the course of several generations as cities continue to grow with the incorporation of climate refugees. Our research will provide data concerning the projected increase in population over the coming two decades in Morocco, and the population will shift from rural areas to urban centers during that period of time. As a result, urban infrastructure will need to be adapted, developed or built to fit the demand of future internal migrations from rural to urban centers in Morocco. This paper will also examine how past experiences of internally displaced people give insight into the challenges faced by future migrants and, beyond the gathering of data, how people react to internal migration. This study employs four different sets of research tools. First, a large part of this study is archival, which involves compiling the relevant literature on the topic and its complex history. This step also includes gathering data bout migrations in Morocco from public data sources. Once the datasets are collected, the next part of the project involves populating the attribute fields and preprocessing the data to make it understandable and usable by machine learning algorithms. In tandem with the mathematical interpretation of data and projected migrations, this study benefits from a theoretical understanding of the critical apparatus existing around urban development of the 20th and 21st centuries that give us insight into past infrastructure development and the rationale behind it. Once the data is ready to be analyzed, different machine learning algorithms will be experimented (k-clustering, support vector regression, random forest analysis) and the results compared for visualization of the data. The final computational part of this study involves analyzing the data and determining what we can learn from it. This paper helps us to understand future trends of population movements within and between regions of North Africa, which will have an impact on various sectors such as urban development, food distribution and water purification, not to mention the creation of public policy in the countries of this region. One of the strengths of this project is the multi-pronged and cross-disciplinary methodology to the research question, which enables an interchange of knowledge and experiences to facilitate innovative solutions to this complex problem. Multiple and diverse intersecting viewpoints allow an exchange of methodological models that provide fresh and informed interpretations of otherwise objective data.

Keywords: climate change, machine learning, migration, Morocco, urban development

Procedia PDF Downloads 150

Authors: Ameen Jameel Alawneh

Abstract:

A Mobile ad-hoc network (MANET) is a collection of wireless mobile hosts that dynamically form a temporary network without the aid of a system administrator. It has neither fixed infrastructure nor wireless ad hoc sessions. It inherently reaches several nodes with a single transmission, and each node functions as both a host and a router. The network maybe represented as a set of clusters each managed by clusterhead. The cluster size is not fixed and it depends on the movement of nodes. We proposed a clusterhead size algorithm (CHSize). This clustering algorithm can be used by several routing algorithms for ad hoc networks. An elected clusterhead is assigned for communication with all other clusters. Analysis and simulation of the algorithm has been implemented using GloMoSim networks simulator, MATLAB and MAPL11 proved that the proposed algorithm achieves the goals.

Keywords: simulation, MANET, Ad-hoc, cluster head size, linked cluster algorithm, loss and dropped packets

Procedia PDF Downloads 392

Authors: Modesta E. Ezema, Chikwendu V. Alabekee, Victoria N. Ishiwu, Ifeyinwa NwosuArize, Chinedu I. Nwoye

Abstract:

The advent of wireless networking in computing technology cannot be overemphasized, it opened up easy accessibility to information resources, networking made easier and brought internet accessibility to our doorsteps, but despite all these, some mishap came in with it that is causing mayhem in today ‘s overall information security. The cyber criminals will always compromise the integrity of a message that is not encrypted or that is encrypted with a weak algorithm.In other to correct the mayhem, this study focuses on cryptosystem and cryptography. This ensures end to end crypt messaging. The study of various cryptographic algorithms, as well as the techniques and applications of the cryptography for efficiency, were all considered in the work., present and future applications of cryptography were dealt with as well as Quantum Cryptography was exposed as the current and the future area in the development of cryptography. An empirical study was conducted to collect data from network users.

Keywords: algorithm, cryptography, cryptosystem, network

Procedia PDF Downloads 349

Authors: Jorge Maestre Vidal, Ana Lucila Sandoval Orozco, Luis Javier García Villalba

Abstract:

With the increase in popularity of mobile devices, new and varied forms of malware have emerged. Consequently, the organizations for cyberdefense have echoed the need to deploy more effective defensive schemes adapted to the challenges posed by these recent monitoring environments. In order to contribute to their development, this paper presents a malware detection strategy for mobile devices based on sequence alignment algorithms. Unlike the previous proposals, only the system calls performed during the startup of applications are studied. In this way, it is possible to efficiently study in depth, the sequences of system calls executed by the applications just downloaded from app stores, and initialize them in a secure and isolated environment. As demonstrated in the performed experimentation, most of the analyzed malicious activities were successfully identified in their boot processes.

Keywords: android, information security, intrusion detection systems, malware, mobile devices

Procedia PDF Downloads 304

Authors: Richard Ren

Abstract:

Air pollution is a serious danger to international well-being and economies - it will kill an estimated 7 million people every year, costing world economies $2.6 trillion by 2060 due to sick days, healthcare costs, and reduced productivity. In the United States alone, 60,000 premature deaths are caused by poor air quality. For this reason, there is a crucial need to develop effective methods to forecast air quality, which can mitigate air pollution’s detrimental public health effects and associated costs by helping people plan ahead and avoid exposure. The goal of this study is to propose an artificial intelligence framework for predicting future air quality based on timing variables (i.e. season, weekday/weekend), future weather forecasts, as well as past pollutant and air quality measurements. The proposed framework utilizes multiple machine learning algorithms (logistic regression, random forest, neural network) with different specifications and averages the results of the three top-performing models to eliminate inaccuracies, weaknesses, and biases from any one individual model. Over time, the proposed framework uses new data to self-adjust model parameters and increase prediction accuracy. To demonstrate its applicability, a prototype of this framework was created to forecast air quality in Los Angeles, California using datasets from the RP4 weather data repository and EPA pollutant measurement data. The results showed good agreement between the framework’s predictions and real-life observations, with an overall 92% model accuracy. The combined model is able to predict more accurately than any of the individual models, and it is able to reliably forecast season-based variations in air quality levels. Top air quality predictor variables were identified through the measurement of mean decrease in accuracy. This study proposed and demonstrated the efficacy of a comprehensive air quality prediction framework leveraging multiple machine learning algorithms to overcome individual algorithm shortcomings. Future enhancements should focus on expanding and testing a greater variety of modeling techniques within the proposed framework, testing the framework in different locations, and developing a platform to automatically publish future predictions in the form of a web or mobile application. Accurate predictions from this artificial intelligence framework can in turn be used to save and improve lives by allowing individuals to protect their health and allowing governments to implement effective pollution control measures.Air pollution is a serious danger to international wellbeing and economies - it will kill an estimated 7 million people every year, costing world economies $2.6 trillion by 2060 due to sick days, healthcare costs, and reduced productivity. In the United States alone, 60,000 premature deaths are caused by poor air quality. For this reason, there is a crucial need to develop effective methods to forecast air quality, which can mitigate air pollution’s detrimental public health effects and associated costs by helping people plan ahead and avoid exposure. The goal of this study is to propose an artificial intelligence framework for predicting future air quality based on timing variables (i.e. season, weekday/weekend), future weather forecasts, as well as past pollutant and air quality measurements. The proposed framework utilizes multiple machine learning algorithms (logistic regression, random forest, neural network) with different specifications and averages the results of the three top-performing models to eliminate inaccuracies, weaknesses, and biases from any one individual model. Over time, the proposed framework uses new data to self-adjust model parameters and increase prediction accuracy. To demonstrate its applicability, a prototype of this framework was created to forecast air quality in Los Angeles, California using datasets from the RP4 weather data repository and EPA pollutant measurement data. The results showed good agreement between the framework’s predictions and real-life observations, with an overall 92% model accuracy. The combined model is able to predict more accurately than any of the individual models, and it is able to reliably forecast season-based variations in air quality levels. Top air quality predictor variables were identified through the measurement of mean decrease in accuracy. This study proposed and demonstrated the efficacy of a comprehensive air quality prediction framework leveraging multiple machine learning algorithms to overcome individual algorithm shortcomings. Future enhancements should focus on expanding and testing a greater variety of modeling techniques within the proposed framework, testing the framework in different locations, and developing a platform to automatically publish future predictions in the form of a web or mobile application. Accurate predictions from this artificial intelligence framework can in turn be used to save and improve lives by allowing individuals to protect their health and allowing governments to implement effective pollution control measures.Air pollution is a serious danger to international wellbeing and economies - it will kill an estimated 7 million people every year, costing world economies $2.6 trillion by 2060 due to sick days, healthcare costs, and reduced productivity. In the United States alone, 60,000 premature deaths are caused by poor air quality. For this reason, there is a crucial need to develop effective methods to forecast air quality, which can mitigate air pollution’s detrimental public health effects and associated costs by helping people plan ahead and avoid exposure. The goal of this study is to propose an artificial intelligence framework for predicting future air quality based on timing variables (i.e. season, weekday/weekend), future weather forecasts, as well as past pollutant and air quality measurements. The proposed framework utilizes multiple machine learning algorithms (logistic regression, random forest, neural network) with different specifications and averages the results of the three top-performing models to eliminate inaccuracies, weaknesses, and biases from any one individual model. Over time, the proposed framework uses new data to self-adjust model parameters and increase prediction accuracy. To demonstrate its applicability, a prototype of this framework was created to forecast air quality in Los Angeles, California using datasets from the RP4 weather data repository and EPA pollutant measurement data. The results showed good agreement between the framework’s predictions and real-life observations, with an overall 92% model accuracy. The combined model is able to predict more accurately than any of the individual models, and it is able to reliably forecast season-based variations in air quality levels. Top air quality predictor variables were identified through the measurement of mean decrease in accuracy. This study proposed and demonstrated the efficacy of a comprehensive air quality prediction framework leveraging multiple machine learning algorithms to overcome individual algorithm shortcomings. Future enhancements should focus on expanding and testing a greater variety of modeling techniques within the proposed framework, testing the framework in different locations, and developing a platform to automatically publish future predictions in the form of a web or mobile application. Accurate predictions from this artificial intelligence framework can in turn be used to save and improve lives by allowing individuals to protect their health and allowing governments to implement effective pollution control measures.

Keywords: air quality prediction, air pollution, artificial intelligence, machine learning algorithms

Procedia PDF Downloads 127

Authors: Stephen Kirkup

Abstract:

This paper discusses the implementation of the boundary element method (BEM) on an Excel spreadsheet and how it can be used in teaching vector calculus and simulation. There are two separate spreadheets, within which Laplace equation is solved by the BEM in two dimensions (LIBEM2) and axisymmetric three dimensions (LBEMA). The main algorithms are implemented in the associated programming language within Excel, Visual Basic for Applications (VBA). The BEM only requires a boundary mesh and hence it is a relatively accessible method. The BEM in the open spreadsheet environment is demonstrated as being useful as an aid to teaching and learning. The application of the BEM implemented on a spreadsheet for educational purposes in introductory vector calculus and simulation is explored. The development of assignment work is discussed, and sample results from student work are given. The spreadsheets were found to be useful tools in developing the students’ understanding of vector calculus and in simulating heat conduction.

Keywords: boundary element method, Laplace’s equation, vector calculus, simulation, education

Procedia PDF Downloads 163

Authors: Martin K. Steiger, Lukas Heisler, Hans-Georg Brachtendorf

Abstract:

Deep neural networks and their variants form the backbone of many AI applications. Based on the so-called residual networks, a continuous formulation of such models as ordinary differential equations (ODEs) has proven advantageous since different techniques may be applied that significantly increase the learning speed and enable controlled trade-offs with the resulting error at the same time. For the evaluation of such models, high-performance numerical differential equation solvers are used, which also provide the gradients required for training. However, whether classical gradient-based methods are even applicable or which one yields the best results has not been discussed yet. This paper aims to redeem this situation by providing empirical results for different applications.

Keywords: deep neural networks, gradient-based learning, image processing, ordinary differential equation networks

Procedia PDF Downloads 168

Authors: Houda Abadlia, Nadia Smairi, Khaled Ghedira

Abstract:

Multiobjective Particle Swarm Optimization (MOPSO) has shown an effective performance for solving test functions and real-world optimization problems. However, this method has a premature convergence problem, which may lead to lack of diversity. In order to improve its performance, this paper presents a hybrid approach which embedded the MOPSO into the island model and integrated a local search technique, Variable Neighborhood Search, to enhance the diversity into the swarm. Experiments on two series of test functions have shown the effectiveness of the proposed approach. A comparison with other evolutionary algorithms shows that the proposed approach presented a good performance in solving multiobjective optimization problems.

Keywords: particle swarm optimization, migration, variable neighborhood search, multiobjective optimization

Procedia PDF Downloads 167

Authors: Ewedafe Simon Uzezi

Abstract:

In this paper we presented a method based on Domain Decomposition (DD) for parallelization of 1-D Sugarcane Equation on parallel platform with parallel paradigms on Master-Slave platform using Message Passing Interface (MPI). The 1-D Sugarcane Equation was discretized using explicit method of discretization requiring evaluation nof temporal and spatial distribution of temperature. This platform gives better predictions of the effects of temperature distribution of the sugarcane problem. This work presented parallel overheads with overlapping communication and communication across parallel computers with numerical results across different block sizes with scalability. However, performance improvement strategies from the DD on various mesh sizes were compared experimentally and parallel results show speedup and efficiency for the parallel algorithms design.

Keywords: sugarcane, parallelization, explicit method, domain decomposition, MPI

Procedia PDF Downloads 21

Authors: Tanmay Raj, Vedika Gupta

Abstract:

For its groundbreaking and substantial power, cloud computing is today’s most popular breakthrough. It is a sort of Internet-based computing that allows users to request and receive numerous services in a cost-effective manner. Virtualization, grid computing, and utility computing are the most widely employed emerging technologies in cloud computing, making it the most powerful. However, cloud computing still has a number of key challenges, such as security, load balancing, and non-critical failure adaption, to name a few. The massive growth of cloud computing will put an undue strain on servers. As a result, network performance will deteriorate. A good load balancing adjustment can make cloud computing more productive and in- crease client fulfillment execution. Load balancing is an important part of cloud computing because it prevents certain nodes from being overwhelmed while others are idle or have little work to perform. Response time, cost, throughput, performance, and resource usage are all parameters that may be improved using load balancing.

Keywords: cloud computing, load balancing, computing, SIP server clusters

Procedia PDF Downloads 124

Authors: Meruyert Serik, Nassipzhan Duisegaliyeva, Danara Tleumagambetova, Madina Ermaganbetova

Abstract:

This research paper explores the process of creating a proctoring system by evaluating the implementation of practical face recognition algorithms. Students of educational programs reviewed the research work "6B01511-Computer Science", "7M01511-Computer Science", "7M01525- STEM Education," and "8D01511-Computer Science" of Eurasian National University named after L.N. Gumilyov. As an outcome, a proctoring system will be created, enabling the conduction of tests and ensuring academic integrity checks within the system. Due to the correct operation of the system, test works are carried out. The result of the creation of the proctoring system will be the basis for the automation of the informational, educational portal developed by machine learning.

Keywords: artificial intelligence, education portal, face recognition, machine learning, proctoring

Procedia PDF Downloads 126

Authors: Conrad Weber

Abstract:

The modelling of railway noise over large projects areas can be very time consuming in terms of preparing the noise models and calculation time. An open-source GIS program has been utilised to assist with the modelling of operational noise levels for 675km of railway corridor. A range of GIS algorithms were utilised to break up the noise model area into manageable calculation sizes. GIS was utilised to prepare and filter a range of noise modelling inputs, including building files, land uses and ground terrain. A spreadsheet was utilised to manage the accuracy of key input parameters, including train speeds, train types, curve corrections, bridge corrections and engine notch settings. GIS was utilised to present the final noise modelling results. This paper explains the noise modelling process and how the spreadsheet and GIS were utilised to accurately model this massive project efficiently.

Keywords: noise, modeling, GIS, rail

Procedia PDF Downloads 122

Authors: Ijaz Ahmed, Anwar Zada, Muhammad Waqas, M. U. Ashraf

Abstract:

We demonstrate the performance of a very efficient tagger applies on hadronically decaying top quark pairs as signal based on deep neural network algorithms and compares with the QCD multi-jet background events. A significant enhancement of performance in boosted top quark events is observed with our limited computing resources. We also compare modern machine learning approaches and perform a multivariate analysis of boosted top-pair as well as single top quark production through weak interaction at √s = 14 TeV proton-proton Collider. The most relevant known background processes are incorporated. Through the techniques of Boosted Decision Tree (BDT), likelihood and Multlayer Perceptron (MLP) the analysis is trained to observe the performance in comparison with the conventional cut based and count approach

Keywords: top tagger, multivariate, deep learning, LHC, single top

Procedia PDF Downloads 111

Authors: M. F. F. Ab Rashid, A. N. Mohd Rose, N. M. Z. Nik Mohamed, W. S. Wan Harun, S. A. Che Ghani

Abstract:

Traveling salesman problem with precedence constraint (TSPPC) is one of the most complex problems in combinatorial optimization. The existing algorithms to solve TSPPC cost large computational time to find the optimal solution. The purpose of this paper is to present an efficient genetic algorithm that guarantees optimal solution with less number of generations and iterations time. Unlike the existing algorithm that generates priority factor as chromosome, the proposed algorithm directly generates sequence of solution as chromosome. As a result, the proposed algorithm is capable of generating optimal solution with smaller number of generations and iteration time compare to existing algorithm.

Keywords: traveling salesman problem, sequencing, genetic algorithm, precedence constraint

Procedia PDF Downloads 560