Search results for: optimization for learning and data analysis

Authors: Pham Van Nha, Le Cam Binh

Abstract:

Particle swarm optimization (PSO) is a population-based stochastic optimization algorithm. PSO was inspired by the natural behavior of birds and fish in migration and foraging for food. PSO is considered as a multidisciplinary optimization model that can be applied in various optimization problems. PSO’s ideas are simple and easy to understand but PSO is only applied in simple model problems. We think that in order to expand the applicability of PSO in complex problems, PSO should be described more explicitly in the form of a mathematical model. In this paper, we represent PSO in a mathematical model and apply in the multivariate data classification. First, PSOs general mathematical model (MPSO) is analyzed as a universal optimization model. Then, Model of Optimal Centroids (MOC) is proposed for the multivariate data classification. Experiments were conducted on some benchmark data sets to prove the effectiveness of MOC compared with several proposed schemes.

Keywords: Analysis of optimization, artificial intelligence-based optimization, optimization for learning and data analysis, global optimization.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 826

Authors: M. R. Aghaebrahimi, S. H. Zahiri, M. Amiri

Abstract:

In this paper a data miner based on the learning automata is proposed and is called LA-miner. The LA-miner extracts classification rules from data sets automatically. The proposed algorithm is established based on the function optimization using learning automata. The experimental results on three benchmarks indicate that the performance of the proposed LA-miner is comparable with (sometimes better than) the Ant-miner (a data miner algorithm based on the Ant Colony optimization algorithm) and CNZ (a well-known data mining algorithm for classification).

Keywords: Data mining, Learning automata, Classification rules, Knowledge discovery.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1897

Authors: William Anderson, Kyle Kobold, Oleg Yakimenko

Abstract:

Green microgrids using mostly renewable energy (RE) for generation, are complex systems with inherent nonlinear dynamics. Among a variety of different optimization tools there are only a few ones that adequately consider this complexity. This paper evaluates applicability of two somewhat similar optimization tools tailored for standalone RE microgrids and also assesses a machine learning tool for performance prediction that can enhance the reliability of any chosen optimization tool. It shows that one of these microgrid optimization tools has certain advantages over another and presents a detailed routine of preparing input data to simulate RE microgrid behavior. The paper also shows how neural-network-based predictive modeling can be used to validate and forecast solar power generation based on weather time series data, which improves the overall quality of standalone RE microgrid analysis.

Keywords: Microgrid, renewable energy, complex systems, optimization, predictive modeling, neural networks.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1000

Authors: H. Bouhamed, A. Rebai, T. Lecroq, M. Jaoua

Abstract:

The objective of our work is to develop a new approach for discovering knowledge from a large mass of data, the result of applying this approach will be an expert system that will serve as diagnostic tools of a phenomenon related to a huge information system. We first recall the general problem of learning Bayesian network structure from data and suggest a solution for optimizing the complexity by using organizational and optimization methods of data. Afterward we proposed a new heuristic of learning a Multi-Entities Bayesian Networks structures. We have applied our approach to biological facts concerning hereditary complex illnesses where the literatures in biology identify the responsible variables for those diseases. Finally we conclude on the limits arched by this work.

Keywords: Data-organization, data-optimization, automatic knowledge discovery, Multi-Entities Bayesian networks, score merging.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1571

Authors: C. Ramsauer, J. Karandikar, D. Leitner, T. Schmitz, F. Bleicher

Abstract:

Machining instability, or chatter, can impose an important limitation to discrete part machining. In this work, a networked implementation of milling stability optimization with Bayesian learning is presented. The milling process was monitored with a wireless sensory tool holder instrumented with an accelerometer at the TU Wien, Vienna, Austria. The recorded data from a milling test cut were used to classify the cut as stable or unstable based on a frequency analysis. The test cut result was used in a Bayesian stability learning algorithm at the University of Tennessee, Knoxville, Tennessee, USA. The algorithm calculated the probability of stability as a function of axial depth of cut and spindle speed based on the test result and recommended parameters for the next test cut. The iterative process between two transatlantic locations was repeated until convergence to a stable optimal process parameter set was achieved.

Keywords: Bayesian learning, instrumented tool holder, machining stability, optimization strategy.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 447

Authors: Vincenzo Capalbo, Gianpaolo Ghiani, Emanuele Manni

Abstract:

Global e-commerce sales have reached unprecedented levels in the past few years. As this trend is only predicted to go up as we continue into the ’20s, new challenges will be faced by companies when planning and controlling e-commerce logistics. In this paper, we survey the related literature on Optimization and Machine Learning as well as on combined methodologies. We also identify the distinctive features of next-generation planning algorithms - namely scalability, model-and-run features and learning capabilities - that will be fundamental to cope with the scale and complexity of logistics in the next decade.

Keywords: e-Commerce, Logistics, Machine Learning, Optimization.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1045

Authors: Wang Lin, Li Zhiqiang

Abstract:

The purpose of this paper is to analyze the cooperative learning behavior pattern based on the data of students' movement. The study firstly reviewed the cooperative learning theory and its research status, and briefly introduced the k-means clustering algorithm. Then, it used clustering algorithm and mathematical statistics theory to analyze the activity rhythm of individual student and groups in different functional areas, according to the movement data provided by 10 first-year graduate students. It also focused on the analysis of students' behavior in the learning area and explored the law of cooperative learning behavior. The research result showed that the cooperative learning behavior analysis method based on movement data proposed in this paper is feasible. From the results of data analysis, the characteristics of behavior of students and their cooperative learning behavior patterns could be found.

Keywords: Behavior pattern, cooperative learning, data analyze, K-means clustering algorithm.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 759

Authors: T. Niknam, M. Nayeripour, B.Bahmani Firouzi

Abstract:

Clustering techniques have received attention in many areas including engineering, medicine, biology and data mining. The purpose of clustering is to group together data points, which are close to one another. The K-means algorithm is one of the most widely used techniques for clustering. However, K-means has two shortcomings: dependency on the initial state and convergence to local optima and global solutions of large problems cannot found with reasonable amount of computation effort. In order to overcome local optima problem lots of studies done in clustering. This paper is presented an efficient hybrid evolutionary optimization algorithm based on combining Particle Swarm Optimization (PSO) and Ant Colony Optimization (ACO), called PSO-ACO, for optimally clustering N object into K clusters. The new PSO-ACO algorithm is tested on several data sets, and its performance is compared with those of ACO, PSO and K-means clustering. The simulation results show that the proposed evolutionary optimization algorithm is robust and suitable for handing data clustering.

Keywords: Ant Colony Optimization (ACO), Data clustering, Hybrid evolutionary optimization algorithm, K-means clustering, Particle Swarm Optimization (PSO).

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2155

Authors: Romisaa Ali

Abstract:

The Policy Gradient approach is a subset of the Deep Reinforcement Learning (DRL) combines Deep Neural Networks (DNN) with Reinforcement Learning (RL). This approach finds the optimal policy of robot movement, based on the experience it gains from interaction with its environment. Unlike previous policy gradient algorithms, which were unable to handle the two types of error variance and bias introduced by the DNN model due to over- or underestimation, this algorithm is capable of handling both types of error variance and bias. This article will discuss the state-of-the-art SOTA policy gradient technique, trust region policy optimization (TRPO), by applying this method in various environments compared to another policy gradient method, the Proximal Policy Optimization (PPO), to explain their robust optimization, using this SOTA to gather experience data during various training phases after observing the impact of hyper-parameters on neural network performance.

Keywords: Deep neural networks, deep reinforcement learning, Proximal Policy Optimization, state-of-the-art, trust region policy optimization.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 111

Authors: M. A. Rubio, A. Urquia

Abstract:

Genetic algorithms (GA) are applied to the solution of high-dimensional optimization problems. Additionally, sensitivity analysis (SA) is usually carried out to determine the effect on optimal solutions of changes in parameter values of the objective function. These two analyses (i.e., optimization and sensitivity analysis) are computationally intensive when applied to high-dimensional functions. The approach presented in this paper consists in performing the SA during the GA execution, by statistically analyzing the data obtained of running the GA. The advantage is that in this case SA does not involve making additional evaluations of the objective function and, consequently, this proposed approach requires less computational effort than conducting optimization and SA in two consecutive steps.

Keywords: Optimization, sensitivity, genetic algorithms, model calibration.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1423

Authors: Yahya H. Zweiri

Abstract:

The back-propagation algorithm calculates the weight changes of an artificial neural network, and a two-term algorithm with a dynamically optimal learning rate and a momentum factor is commonly used. Recently the addition of an extra term, called a proportional factor (PF), to the two-term BP algorithm was proposed. The third term increases the speed of the BP algorithm. However, the PF term also reduces the convergence of the BP algorithm, and optimization approaches for evaluating the learning parameters are required to facilitate the application of the three terms BP algorithm. This paper considers the optimization of the new back-propagation algorithm by using derivative information. A family of approaches exploiting the derivatives with respect to the learning rate, momentum factor and proportional factor is presented. These autonomously compute the derivatives in the weight space, by using information gathered from the forward and backward procedures. The three-term BP algorithm and the optimization approaches are evaluated using the benchmark XOR problem.

Keywords: Neural Networks, Backpropagation, Optimization.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1500

Authors: B.R. Potgieter, G. Venter

Abstract:

Numerical design optimization is a powerful tool that can be used by engineers during any stage of the design process. There are many different applications for structural optimization. A specific application that will be discussed in the following paper is experimental data matching. Data obtained through tests on a physical structure will be matched with data from a numerical model of that same structure. The data of interest will be the dynamic characteristics of an antenna structure focusing on the mode shapes and modal frequencies. The structure used was a scaled and simplified model of the Karoo Array Telescope-7 (KAT-7) antenna structure. This kind of data matching is a complex and difficult task. This paper discusses how optimization can assist an engineer during the process of correlating a finite element model with vibration test data.

Keywords: Finite Element Model (FEM), Karoo Array Telescope(KAT-7), modal frequencies, mode shapes, optimization, shape optimization, size optimization, vibration tests

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1798

Authors: Arda Yunianta, Norazah Yusof, Mohd Shahizan Othman, Dewi Octaviani

Abstract:

Learning is the acquisition of new mental schemata, knowledge, abilities and skills which can be used to solve problems potentially more successfully. The learning process is optimum when it is assisted and personalized. Learning is not a single activity, but should involve many possible activities to make learning become meaningful. Many e-learning applications provide facilities to support teaching and learning activities. One way to identify whether the e-learning system is being used by the learners is through the number of hits that can be obtained from the e-learning system's log data. However, we cannot rely solely to the number of hits in order to determine whether learning had occurred meaningfully. This is due to the fact that meaningful learning should engage five characteristics namely active, constructive, intentional, authentic and cooperative. This paper aims to analyze the e-learning activities that is meaningful to learning. By focusing on the meaningful learning characteristics, we match it to the corresponding Moodle e-learning activities. This analysis discovers the activities that have high impact to meaningful learning, as well as activities that are less meaningful. The high impact activities is given high weights since it become important to meaningful learning, while the low impact has less weight and said to be supportive e-learning activities. The result of this analysis helps us categorize which e-learning activities that are meaningful to learning and guide us to measure the effectiveness of e-learning usage.

Keywords: e-learning system, e-learning activity, meaningful learning characteristics, Moodle

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 3090

Authors: Varun Singh, Mainak Bandyopadhyay, Maharana Pratap Singh

Abstract:

The shortest path (SP) problem concerns with finding the shortest path from a specific origin to a specified destination in a given network while minimizing the total cost associated with the path. This problem has widespread applications. Important applications of the SP problem include vehicle routing in transportation systems particularly in the field of in-vehicle Route Guidance System (RGS) and traffic assignment problem (in transportation planning). Well known applications of evolutionary methods like Genetic Algorithms (GA), Ant Colony Optimization, Particle Swarm Optimization (PSO) have come up to solve complex optimization problems to overcome the shortcomings of existing shortest path analysis methods. It has been reported by various researchers that PSO performs better than other evolutionary optimization algorithms in terms of success rate and solution quality. Further Geographic Information Systems (GIS) have emerged as key information systems for geospatial data analysis and visualization. This research paper is focused towards the application of PSO for solving the shortest path problem between multiple points of interest (POI) based on spatial data of Allahabad City and traffic speed data collected using GPS. Geovisualization of results of analysis is carried out in GIS.

Keywords: GIS, Outliers, PSO, Traffic Data.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2829

Authors: Parvaneh Shabanzadeh, Malik Hj Abu Hassan, Leong Wah June, Maryam Mohagheghtabar

Abstract:

Clustering is one of an interesting data mining topics that can be applied in many fields. Recently, the problem of cluster analysis is formulated as a problem of nonsmooth, nonconvex optimization, and an algorithm for solving the cluster analysis problem based on nonsmooth optimization techniques is developed. This optimization problem has a number of characteristics that make it challenging: it has many local minimum, the optimization variables can be either continuous or categorical, and there are no exact analytical derivatives. In this study we show how to apply a particular class of optimization methods known as pattern search methods to address these challenges. These methods do not explicitly use derivatives, an important feature that has not been addressed in previous studies. Results of numerical experiments are presented which demonstrate the effectiveness of the proposed method.

Keywords: Clustering functions, Non-smooth Optimization, Nonconvex Optimization, Pattern Search Method.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1585

Authors: Ekachai Phaisangittisagul, Rapeepol Chongprachawat

Abstract:

Obtaining labeled data in supervised learning is often difficult and expensive, and thus the trained learning algorithm tends to be overfitting due to small number of training data. As a result, some researchers have focused on using unlabeled data which may not necessary to follow the same generative distribution as the labeled data to construct a high-level feature for improving performance on supervised learning tasks. In this paper, we investigate the impact of the relationship between unlabeled and labeled data for classification performance. Specifically, we will apply difference unlabeled data which have different degrees of relation to the labeled data for handwritten digit classification task based on MNIST dataset. Our experimental results show that the higher the degree of relation between unlabeled and labeled data, the better the classification performance. Although the unlabeled data that is completely from different generative distribution to the labeled data provides the lowest classification performance, we still achieve high classification performance. This leads to expanding the applicability of the supervised learning algorithms using unsupervised learning.

Keywords: Autoencoder, high-level feature, MNIST dataset, selftaught learning, supervised learning.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1775

Authors: David W.S. Tai, Ren-Cheng Zhang, Sheng-Hung Chang, Chin-Pin Chen, Jia-Ling Chen

Abstract:

This study reports results of a meta-analytic path analysis e-learning Acceptance Model with k = 27 studies, Databases searched included Information Sciences Institute (ISI) website. Variables recorded included perceived usefulness, perceived ease of use, attitude toward behavior, and behavioral intention to use e-learning. A correlation matrix of these variables was derived from meta-analytic data and then analyzed by using structural path analysis to test the fitness of the e-learning acceptance model to the observed aggregated data. Results showed the revised hypothesized model to be a reasonable, good fit to aggregated data. Furthermore, discussions and implications are given in this article.

Keywords: E-learning, Meta Analytic Path Analysis, Technology Acceptance Model

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2395

Authors: M. Sadeghzadeh, M. Teshnehlab

Abstract:

Feature selection has recently been the subject of intensive research in data mining, specially for datasets with a large number of attributes. Recent work has shown that feature selection can have a positive effect on the performance of machine learning algorithms. The success of many learning algorithms in their attempts to construct models of data, hinges on the reliable identification of a small set of highly predictive attributes. The inclusion of irrelevant, redundant and noisy attributes in the model building process phase can result in poor predictive performance and increased computation. In this paper, a novel feature search procedure that utilizes the Ant Colony Optimization (ACO) is presented. The ACO is a metaheuristic inspired by the behavior of real ants in their search for the shortest paths to food sources. It looks for optimal solutions by considering both local heuristics and previous knowledge. When applied to two different classification problems, the proposed algorithm achieved very promising results.

Keywords: Ant colony optimization, Classification, Datamining, Feature selection.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2374

Authors: [email protected]

Abstract:

Breast cancer is one of the most common types in women. Early prediction of breast cancer helps physicians detect cancer in its early stages. Big cancer data need a very powerful tool to analyze and extract predictions. Machine learning and deep learning are two of the most efficient tools for predicting cancer based on textual data. In this study, we developed a fusion model of two machine learning and deep learning models. To obtain the final prediction, Long-Short Term Memory (LSTM), ensemble learning with hyper parameters optimization, and score-level fusion is used. Experiments are done on the Breast Cancer Surveillance Consortium (BCSC) dataset after balancing and grouping the class categories. Five different training scenarios are used, and the tests show that the designed fusion model improved the performance by 3.3% compared to the individual models.

Keywords: Machine learning, Deep learning, cancer prediction, breast cancer, LSTM, Score-Level Fusion.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 303

Authors: M. R. Aghaebrahimi, S. H. Zahiri, M. Amiri

Abstract:

The necessity of solving multi dimensional complicated scientific problems beside the necessity of several objective functions optimization are the most motive reason of born of artificial intelligence and heuristic methods. In this paper, we introduce a new method for multiobjective optimization based on learning automata. In the proposed method, search space divides into separate hyper-cubes and each cube is considered as an action. After gathering of all objective functions with separate weights, the cumulative function is considered as the fitness function. By the application of all the cubes to the cumulative function, we calculate the amount of amplification of each action and the algorithm continues its way to find the best solutions. In this Method, a lateral memory is used to gather the significant points of each iteration of the algorithm. Finally, by considering the domination factor, pareto front is estimated. Results of several experiments show the effectiveness of this method in comparison with genetic algorithm based method.

Keywords: Function optimization, Multiobjective optimization, Learning automata.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1639

Authors: Ioannis P. Panapakidis, Marios N. Moschakis

Abstract:

The Greek Energy Market is structured as a mandatory pool where the producers make their bid offers in day-ahead basis. The System Operator solves an optimization routine aiming at the minimization of the cost of produced electricity. The solution of the optimization problem leads to the calculation of the System Marginal Price (SMP). Accurate forecasts of the SMP can lead to increased profits and more efficient portfolio management from the producer`s perspective. Aim of this study is to provide a comparative analysis of various machine learning models such as artificial neural networks and neuro-fuzzy models for the prediction of the SMP of the Greek market. Machine learning algorithms are favored in predictions problems since they can capture and simulate the volatilities of complex time series.

Keywords: Deregulated energy market, forecasting, machine learning, system marginal price, energy efficiency and quality.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1244

Authors: M. Almohaimeed, Y. Alhaidari, H. Alhamdan, A. Alfaries, A. Ater Kranov

Abstract:

A university-wide survey to obtain baseline data regarding the perceptions of key terms related to e-learning and distance learning among students, faculty and staff was conducted to help achieve the goals of Princess Nourah bint Abdulrahman University’s and the Kingdom of Saudi Arabia’s National Center for e-learning and Distance Learning. This paper comprises a relevant literature review, the survey methodology, preliminary data analysis, discussion, and recommendations for further research. The major findings indicate a deep and wide differentiation of understanding among users of critical key terms.

Keywords: E-learning, distance learning, on-line learning, perceptions of learning environments.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1774

Authors: Liang Zhao, Jili Zhang, Chongquan Zhong

Abstract:

Energy consumption data, in particular those involving public buildings, are impacted by many factors: the building structure, climate/environmental parameters, construction, system operating condition, and user behavior patterns. Traditional methods for data analysis are insufficient. This paper delves into the data mining technology to determine its application in the analysis of building energy consumption data including energy consumption prediction, fault diagnosis, and optimal operation. Recent literature are reviewed and summarized, the problems faced by data mining technology in the area of energy consumption data analysis are enumerated, and research points for future studies are given.

Keywords: Data mining, data analysis, prediction, optimization, building operational performance.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 3647

Authors: Engin Yesil, Leon Urbas

Abstract:

Modeling of complex dynamic systems, which are very complicated to establish mathematical models, requires new and modern methodologies that will exploit the existing expert knowledge, human experience and historical data. Fuzzy cognitive maps are very suitable, simple, and powerful tools for simulation and analysis of these kinds of dynamic systems. However, human experts are subjective and can handle only relatively simple fuzzy cognitive maps; therefore, there is a need of developing new approaches for an automated generation of fuzzy cognitive maps using historical data. In this study, a new learning algorithm, which is called Big Bang-Big Crunch, is proposed for the first time in literature for an automated generation of fuzzy cognitive maps from data. Two real-world examples; namely a process control system and radiation therapy process, and one synthetic model are used to emphasize the effectiveness and usefulness of the proposed methodology.

Keywords: Big Bang-Big Crunch optimization, Dynamic Systems, Fuzzy Cognitive Maps, Learning.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1800

Authors: Wentian Shi, Daming Shi, Maysam Orouskhani, Feng Tian

Abstract:

Currently the most prevalent deep learning methods require a large amount of data for training, whereas few-shot learning tries to learn a model from limited data without extensive retraining. In this paper, we present a loss function based on triplet loss for solving few-shot problem using metric based learning. Instead of setting the margin distance in triplet loss as a constant number empirically, we propose an adaptive margin distance strategy to obtain the appropriate margin distance automatically. We implement the strategy in the deep siamese network for deep metric embedding, by utilizing an optimization approach by penalizing the worst case and rewarding the best. Our experiments on image recognition and co-segmentation model demonstrate that using our proposed triplet loss with adaptive margin distance can significantly improve the performance.

Keywords: Few-shot learning, triplet network, adaptive margin, deep learning.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 825

Authors: S. Sarumathi, N. Shanthi

Abstract:

Due to the fast and flawless technological innovation there is a tremendous amount of data dumping all over the world in every domain such as Pattern Recognition, Machine Learning, Spatial Data Mining, Image Analysis, Fraudulent Analysis, World Wide Web etc., This issue turns to be more essential for developing several tools for data mining functionalities. The major aim of this paper is to analyze various tools which are used to build a resourceful analytical or descriptive model for handling large amount of information more efficiently and user friendly. In this survey the diverse tools are illustrated with their extensive technical paradigm, outstanding graphical interface and inbuilt multipath algorithms in which it is very useful for handling significant amount of data more indeed.

Keywords: Classification, Clustering, Data Mining, Machine learning, Visualization.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2397

Authors: Shuen-Tai Wang, Fang-An Kuo, Chau-Yi Chou, Yu-Bin Fang

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.

Keywords: Artificial Intelligence, machine learning, deep learning, convolutional neural networks.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1180

Authors: S. Assi, M. Haffar

Abstract:

Teaching statistics is a critical and challenging issue especially to students from multidisciplinary and diverse postgraduate backgrounds. Postgraduate research students require statistics not only for the design of experiments; but also for data analysis. Students often perceive statistics as a complex and technical subject; thus, they leave data analysis to the last moment. The lecture needs to be simple and inclusive at the same time to make it comprehendible and address the learning needs of each student. Therefore, the aim of this work was to design a simple and comprehendible statistics lecture to postgraduate research students regarding ‘Research plan, design and data collection’. The lecture adopted the constructive alignment learning theory which facilitated the learning environments for the students. The learning environment utilized a student-centered approach and used interactive learning environment with in-class discussion, handouts and electronic voting system handsets. For evaluation of the lecture, formative assessment was made with in-class discussions and poll questions which were introduced during and after the lecture. The whole approach showed to be effective in creating a learning environment to the students who were able to apply the concepts addressed to their individual research projects.

Keywords: Teaching, statistics, lecture, multidisciplinary, postgraduate, learning theory, learning environment, student-centered approach, data analysis.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1085

Authors: Michal Kopcek, Tomas Skulavik, Michal Kebisek, Gabriela Krizanova

Abstract:

The operation of nuclear power plants involves continuous monitoring of the environment in their area. This monitoring is performed using a complex data acquisition system, which collects status information about the system itself and values of many important physical variables e.g. temperature, humidity, dose rate etc. This paper describes a proposal and optimization of communication that takes place in teledosimetric system between the central control server responsible for the data processing and storing and the decentralized measuring stations, which are measuring the physical variables. Analyzes of ongoing communication were performed and consequently the optimization of the system architecture and communication was done.

Keywords: Communication protocol, transmission optimization, data acquisition.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1769

Authors: Eyup Yurt, Ahmet Kurnaz, Ismail Sahin

Abstract:

The purpose of the study is to investigate the education faculty students’ attitudes towards e-learning according to different variables. In current study, the data were collected from 393 students of an education faculty in Turkey. In this study, theattitude towards e‐learning scale and the demographic information form were used to collect data. The collected data were analyzed by t-test, ANOVA and Pearson correlation coefficient. It was found that there is a significant difference in students’ tendency towards e-learning and avoidance from e-learning based on gender. Male students have more positive attitudes towards e-learning than female students. Also, the students who used the internet lesshave higher levels of avoidance from e-learning. Additionally, it is found that there is a positive and significant relationship between the number of personal mobile learning devices and tendency towards e-learning. On the other hand, there is a negative and significant relationship between the number of personal mobile learning devices and avoidance from e-learning. Also, suggestions were presented according to findings.

Keywords: Education faculty students, attitude towards e-learning, gender, daily Internet usage time, m-learning.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2123