Search results for: Network reduction

Authors: H. Shayeghi, H. A. Shayanfar, G. Azimi

Abstract:

Short-Term Load Forecasting (STLF) plays an important role for the economic and secure operation of power systems. In this paper, Continuous Genetic Algorithm (CGA) is employed to evolve the optimum large neural networks structure and connecting weights for one-day ahead electric load forecasting problem. This study describes the process of developing three layer feed-forward large neural networks for load forecasting and then presents a heuristic search algorithm for performing an important task of this process, i.e. optimal networks structure design. The proposed method is applied to STLF of the local utility. Data are clustered due to the differences in their characteristics. Special days are extracted from the normal training sets and handled separately. In this way, a solution is provided for all load types, including working days and weekends and special days. We find good performance for the large neural networks. The proposed methodology gives lower percent errors all the time. Thus, it can be applied to automatically design an optimal load forecaster based on historical data.

Keywords: Feed-forward Large Neural Network, Short-TermLoad Forecasting, Continuous Genetic Algorithm.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1825

Authors: S. K. Tomar, R. Prasad, S. Panda, C. Ardil

Abstract:

Reduction of Single Input Single Output (SISO) discrete systems into lower order model, using a conventional and an evolutionary technique is presented in this paper. In the conventional technique, the mixed advantages of Modified Cauer Form (MCF) and differentiation are used. In this method the original discrete system is, first, converted into equivalent continuous system by applying bilinear transformation. The denominator of the equivalent continuous system and its reciprocal are differentiated successively, the reduced denominator of the desired order is obtained by combining the differentiated polynomials. The numerator is obtained by matching the quotients of MCF. The reduced continuous system is converted back into discrete system using inverse bilinear transformation. In the evolutionary technique method, Particle Swarm Optimization (PSO) is employed to reduce the higher order model. PSO method is based on the minimization of the Integral Squared Error (ISE) between the transient responses of original higher order model and the reduced order model pertaining to a unit step input. Both the methods are illustrated through numerical example.

Keywords: Discrete System, Single Input Single Output (SISO), Bilinear Transformation, Reduced Order Model, Modified CauerForm, Polynomial Differentiation, Particle Swarm Optimization, Integral Squared Error.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1938

Authors: Umber Iqbal, Shaleeza Sohail, Muhammad Younas Javed

Abstract:

True integration of multimedia services over wired or wireless networks increase the productivity and effectiveness in today-s networks. IP Multimedia Subsystems are Next Generation Network architecture to provide the multimedia services over fixed or mobile networks. This paper proposes an extended SIP-based QoS Management architecture for IMS services over underlying IP access networks. To guarantee the end-to-end QoS for IMS services in interconnection backbone, SIP based proxy Modules are introduced to support the QoS provisioning and to reduce the handoff disruption time over IP access networks. In our approach these SIP Modules implement the combination of Diffserv and MPLS QoS mechanisms to assure the guaranteed QoS for real-time multimedia services. To guarantee QoS over access networks, SIP Modules make QoS resource reservations in advance to provide best QoS to IMS users over heterogeneous networks. To obtain more reliable multimedia services, our approach allows the use of SCTP protocol over SIP instead of UDP due to its multi-streaming feature. This architecture enables QoS provisioning for IMS roaming users to differentiate IMS network from other common IP networks for transmission of realtime multimedia services. To validate our approach simulation models are developed on short scale basis. The results show that our approach yields comparable performance for efficient delivery of IMS services over heterogeneous IP access networks.

Keywords: SIP-Based QoS Management Architecture, IPMultimedia Subsystems, IP Access Networks

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2612

Authors: B. Janani, S. Thiruvenkadam

Abstract:

In this work, the system evaluates the impact of considering a stochastic approach on the day ahead basis Unit Commitment. Comparisons between stochastic and deterministic Unit Commitment solutions are provided. The Unit Commitment model consists in the minimization of the total operation costs considering unit’s technical constraints like ramping rates, minimum up and down time. Load shedding and wind power spilling is acceptable, but at inflated operational costs. The evaluation process consists in the calculation of the optimal unit commitment and in verifying the fulfillment of the considered constraints. For the calculation of the optimal unit commitment, an algorithm based on the Benders Decomposition, namely on the Dual Dynamic Programming, was developed. Two approaches were considered on the construction of stochastic solutions. Data related to wind power outputs from two different operational days are considered on the analysis. Stochastic and deterministic solutions are compared based on the actual measured wind power output at the operational day. Through a technique capability of finding representative wind power scenarios and its probabilities, the system can analyze a more detailed process about the expected final operational cost.

Keywords: Benders’ decomposition, network constrained AC unit commitment, stochastic programming, wind power uncertainty.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1307

Authors: Guojian Cheng, Tianshi Liu, Jiaxin Han, Zheng Wang

Abstract:

The competitive learning is an adaptive process in which the neurons in a neural network gradually become sensitive to different input pattern clusters. The basic idea behind the Kohonen-s Self-Organizing Feature Maps (SOFM) is competitive learning. SOFM can generate mappings from high-dimensional signal spaces to lower dimensional topological structures. The main features of this kind of mappings are topology preserving, feature mappings and probability distribution approximation of input patterns. To overcome some limitations of SOFM, e.g., a fixed number of neural units and a topology of fixed dimensionality, Growing Self-Organizing Neural Network (GSONN) can be used. GSONN can change its topological structure during learning. It grows by learning and shrinks by forgetting. To speed up the training and convergence, a new variant of GSONN, twin growing cell structures (TGCS) is presented here. This paper first gives an introduction to competitive learning, SOFM and its variants. Then, we discuss some GSONN with fixed dimensionality, which include growing cell structures, its variants and the author-s model: TGCS. It is ended with some testing results comparison and conclusions.

Keywords: Artificial neural networks, Competitive learning, Growing cell structures, Self-organizing feature maps.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1536

Authors: John Kabuba, Antoine Mulaba-Bafubiandi, Kim Battle

Abstract:

This paper presents a critical study about the application of Neural Networks to ion-exchange process. Ionexchange is a complex non-linear process involving many factors influencing the ions uptake mechanisms from the pregnant solution. The following step includes the elution. Published data presents empirical isotherm equations with definite shortcomings resulting in unreliable predictions. Although Neural Network simulation technique encounters a number of disadvantages including its “black box", and a limited ability to explicitly identify possible causal relationships, it has the advantage to implicitly handle complex nonlinear relationships between dependent and independent variables. In the present paper, the Neural Network model based on the back-propagation algorithm Levenberg-Marquardt was developed using a three layer approach with a tangent sigmoid transfer function (tansig) at hidden layer with 11 neurons and linear transfer function (purelin) at out layer. The above mentioned approach has been used to test the effectiveness in simulating ion exchange processes. The modeling results showed that there is an excellent agreement between the experimental data and the predicted values of copper ions removed from aqueous solutions.

Keywords: Copper, ion-exchange process, neural networks, simulation

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1624

Authors: K. Paradkar, S. N. Mudliar, A. Sharma, A. B. Pandit, R. A. Pandey

Abstract:

The study explores the hybrid combination of Hydrodynamic Cavitation (HC) and Subcritical Wet Air Oxidation-based pretreatment of complex industrial effluent to enhance the biodegradability selectively (without major COD destruction) to facilitate subsequent enhanced downstream processing via anaerobic or aerobic biological treatment. Advanced oxidation based techniques can be less efficient as standalone options and a hybrid approach by combining Hydrodynamic Cavitation (HC), and Wet Air Oxidation (WAO) can lead to a synergistic effect since both the options are based on common free radical mechanism. The HC can be used for initial turbulence and generation of hotspots which can begin the free radical attack and this agitating mixture then can be subjected to less intense WAO since initial heat (to raise the activation energy) can be taken care by HC alone. Lab-scale venturi-based hydrodynamic cavitation and wet air oxidation reactor with biomethanated distillery wastewater (BMDWW) as a model effluent was examined for establishing the proof-of-concept. The results indicated that for a desirable biodegradability index (BOD: COD - BI) enhancement (up to 0.4), the Cavitation (standalone) pretreatment condition was: 5 bar and 88 min reaction time with a COD reduction of 36 % and BI enhancement of up to 0.27 (initial BI - 0.17). The optimum WAO condition (standalone) was: 150oC, 6 bar and 30 minutes with 31% COD reduction and 0.33 BI. The hybrid pretreatment (combined Cavitation + WAO) worked out to be 23.18 min HC (at 5 bar) followed by 30 min WAO at 150oC, 6 bar, at which around 50% COD was retained yielding a BI of 0.55. FTIR & NMR analysis of pretreated effluent indicated dissociation and/or reorientation of complex organic compounds in untreated effluent to simpler organic compounds post-pretreatment.

Keywords: BI, hybrid, hydrodynamic cavitation, wet air oxidation.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1749

Authors: Emna Benmohamed, Hela Ltifi, Mounir Ben Ayed

Abstract:

Bayesian Network (BN) is one of the most efficient classification methods. It is widely used in several fields (i.e., medical diagnostics, risk analysis, bioinformatics research). The BN is defined as a probabilistic graphical model that represents a formalism for reasoning under uncertainty. This classification method has a high-performance rate in the extraction of new knowledge from data. The construction of this model consists of two phases for structure learning and parameter learning. For solving this problem, the K2 algorithm is one of the representative data-driven algorithms, which is based on score and search approach. In addition, the integration of the expert's knowledge in the structure learning process allows the obtainment of the highest accuracy. In this paper, we propose a hybrid approach combining the improvement of the K2 algorithm called K2 algorithm for Parents and Children search (K2PC) and the expert-driven method for learning the structure of BN. The evaluation of the experimental results, using the well-known benchmarks, proves that our K2PC algorithm has better performance in terms of correct structure detection. The real application of our model shows its efficiency in the analysis of the phosphate laundry effluents' impact on the watershed in the Gafsa area (southwestern Tunisia).

Keywords: Classification, Bayesian network; structure learning, K2 algorithm, expert knowledge, surface water analysis.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 500

Authors: Hosein Hamisheh Bahar, Mohammad Hossein Fazel Zarandi, Akbar Esfahanipour

Abstract:

In this paper, a hybrid expert system is developed by using fuzzy genetic network programming with reinforcement learning (GNP-RL). In this system, the frame-based structure of the system uses the trading rules extracted by GNP. These rules are extracted by using technical indices of the stock prices in the training time period. For developing this system, we applied fuzzy node transition and decision making in both processing and judgment nodes of GNP-RL. Consequently, using these method not only did increase the accuracy of node transition and decision making in GNP's nodes, but also extended the GNP's binary signals to ternary trading signals. In the other words, in our proposed Fuzzy GNP-RL model, a No Trade signal is added to conventional Buy or Sell signals. Finally, the obtained rules are used in a frame-based system implemented in Kappa-PC software. This developed trading system has been used to generate trading signals for ten companies listed in Tehran Stock Exchange (TSE). The simulation results in the testing time period shows that the developed system has more favorable performance in comparison with the Buy and Hold strategy.

Keywords: Fuzzy genetic network programming, hybrid expert system, technical trading signal, Tehran stock exchange.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1850

Authors: M. Nikbakht, N. Choupani

Abstract:

In this paper delamination phenomenon in Carbon-Epoxy laminated composite material is investigated numerically. Arcan apparatus and specimen is modeled in ABAQUS finite element software for different loading conditions and crack geometries. The influence of variation of crack geometry on interlaminar fracture stress intensity factor and energy release rate for various mixed mode ratios and pure mode I and II was studied. Also, correction factors for this specimen for different crack length ratios were calculated. The finite element results indicate that for loading angles close to pure mode-II loading, a high ratio of mode-II to mode-I fracture is dominant and there is an opposite trend for loading angles close to pure mode-I loading. It confirms that by varying the loading angle of Arcan specimen pure mode-I, pure mode-II and a wide range of mixed-mode loading conditions can be created and tested. Also, numerical results confirm that the increase of the mode- II loading contribution leads to an increase of fracture resistance in the CF/PEI composite (i.e., a reduction in the total strain energy release rate) and the increase of the crack length leads to a reduction of interlaminar fracture resistance in the CF/PEI composite (i.e., an increase in the total interlaminar strain energy release rate).

Keywords: Fracture Mechanics, Mixed Mode, Arcan Specimen, Finite Element.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1904

Authors: M. Analoui, M. Fadavi Amiri

Abstract:

The design of a pattern classifier includes an attempt to select, among a set of possible features, a minimum subset of weakly correlated features that better discriminate the pattern classes. This is usually a difficult task in practice, normally requiring the application of heuristic knowledge about the specific problem domain. The selection and quality of the features representing each pattern have a considerable bearing on the success of subsequent pattern classification. Feature extraction is the process of deriving new features from the original features in order to reduce the cost of feature measurement, increase classifier efficiency, and allow higher classification accuracy. Many current feature extraction techniques involve linear transformations of the original pattern vectors to new vectors of lower dimensionality. While this is useful for data visualization and increasing classification efficiency, it does not necessarily reduce the number of features that must be measured since each new feature may be a linear combination of all of the features in the original pattern vector. In this paper a new approach is presented to feature extraction in which feature selection, feature extraction, and classifier training are performed simultaneously using a genetic algorithm. In this approach each feature value is first normalized by a linear equation, then scaled by the associated weight prior to training, testing, and classification. A knn classifier is used to evaluate each set of feature weights. The genetic algorithm optimizes a vector of feature weights, which are used to scale the individual features in the original pattern vectors in either a linear or a nonlinear fashion. By this approach, the number of features used in classifying can be finely reduced.

Keywords: Feature reduction, genetic algorithm, pattern classification, nearest neighbor rule classifiers (k-NNR).

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1758

Authors: Ramaswamy Palaniappan, Nai-Jen Huan

Abstract:

Classification of electroencephalogram (EEG) signals extracted during mental tasks is a technique that is actively pursued for Brain Computer Interfaces (BCI) designs. In this paper, we compared the classification performances of univariateautoregressive (AR) and multivariate autoregressive (MAR) models for representing EEG signals that were extracted during different mental tasks. Multilayer Perceptron (MLP) neural network (NN) trained by the backpropagation (BP) algorithm was used to classify these features into the different categories representing the mental tasks. Classification performances were also compared across different mental task combinations and 2 sets of hidden units (HU): 2 to 10 HU in steps of 2 and 20 to 100 HU in steps of 20. Five different mental tasks from 4 subjects were used in the experimental study and combinations of 2 different mental tasks were studied for each subject. Three different feature extraction methods with 6th order were used to extract features from these EEG signals: AR coefficients computed with Burg-s algorithm (ARBG), AR coefficients computed with stepwise least square algorithm (ARLS) and MAR coefficients computed with stepwise least square algorithm. The best results were obtained with 20 to 100 HU using ARBG. It is concluded that i) it is important to choose the suitable mental tasks for different individuals for a successful BCI design, ii) higher HU are more suitable and iii) ARBG is the most suitable feature extraction method.

Keywords: Autoregressive, Brain-Computer Interface, Electroencephalogram, Neural Network.

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

Authors: Jacqueline R. T. Alipo-on, Francesca I. F. Escobar, Myles J. T. Tan, Hezerul Abdul Karim, Nouar AlDahoul

Abstract:

Electrocardiogram (ECG) signal analysis and processing are crucial in the diagnosis of cardiovascular diseases which are considered as one of the leading causes of mortality worldwide. However, the traditional rule-based analysis of large volumes of ECG data is time-consuming, labor-intensive, and prone to human errors. With the advancement of the programming paradigm, algorithms such as machine learning have been increasingly used to perform an analysis on the ECG signals. In this paper, various deep learning algorithms were adapted to classify five classes of heart beat types. The dataset used in this work is the synthetic MIT-Beth Israel Hospital (MIT-BIH) Arrhythmia dataset produced from generative adversarial networks (GANs). Various deep learning models such as ResNet-50 convolutional neural network (CNN), 1-D CNN, and long short-term memory (LSTM) were evaluated and compared. ResNet-50 was found to outperform other models in terms of recall and F1 score using a five-fold average score of 98.88% and 98.87%, respectively. 1-D CNN, on the other hand, was found to have the highest average precision of 98.93%.

Keywords: Heartbeat classification, convolutional neural network, electrocardiogram signals, ECG signals, generative adversarial networks, long short-term memory, LSTM, ResNet-50.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 169

Authors: Lae-Jeong Park

Abstract:

The paper presents a method that utilizes figure-ground color segmentation to extract effective global feature in terms of false positive reduction in the head-shoulder detection. Conventional detectors that rely on local features such as HOG due to real-time operation suffer from false positives. Color cue in an input image provides salient information on a global characteristic which is necessary to alleviate the false positives of the local feature based detectors. An effective approach that uses figure-ground color segmentation has been presented in an effort to reduce the false positives in object detection. In this paper, an extended version of the approach is presented that adopts separate multipart foregrounds instead of a single prior foreground and performs the figure-ground color segmentation with each of the foregrounds. The multipart foregrounds include the parts of the head-shoulder shape and additional auxiliary foregrounds being optimized by a search algorithm. A classifier is constructed with the feature that consists of a set of the multiple resulting segmentations. Experimental results show that the presented method can discriminate more false positive than the single prior shape-based classifier as well as detectors with the local features. The improvement is possible because the presented approach can reduce the false positives that have the same colors in the head and shoulder foregrounds.

Keywords: Pedestrian detection, color segmentation, false positives, feature extraction.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1138

Authors: Wann-Ming Wey

Abstract:

In recent years, in addition to face the external threats such as energy shortages and climate change, traffic congestion and environmental pollution have become anxious problems for many cities. Considering private automobile-oriented urban development had produced many negative environmental and social impacts, the transit-oriented development (TOD) has been considered as a sustainable urban model. TOD encourages public transport combined with friendly walking and cycling environment designs, however, non-motorized modes help improving human health, energy saving, and reducing carbon emissions. Due to environmental changes often affect the planners’ decision-making; this research applies dynamic network process (DNP) which includes the time dependent concept to promoting friendly walking and cycling environmental designs as an advanced planning support system for environment improvements.

This research aims to discuss what kinds of design strategies can improve a friendly walking and cycling environment under TOD. First of all, we collate and analyze environment designing factors by reviewing the relevant literatures as well as divide into three aspects of “safety”, “convenience”, and “amenity” from fifteen environment designing factors. Furthermore, we utilize fuzzy Delphi Technique (FDT) expert questionnaire to filter out the more important designing criteria for the study case. Finally, we utilized DNP expert questionnaire to obtain the weights changes at different time points for each design criterion. Based on the changing trends of each criterion weight, we are able to develop appropriate designing strategies as the reference for planners to allocate resources in a dynamic environment. In order to illustrate the approach we propose in this research, Taipei city as one example has been used as an empirical study, and the results are in depth analyzed to explain the application of our proposed approach.

Keywords: Environment Planning Support Systems, Walking and Cycling, Transit-oriented Development (TOD), Dynamic Network Process (DNP).

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1844

Authors: M. Moslehpour, S. Khorsandi

Abstract:

Due to shortage in IPv4 addresses, transition to IPv6 has gained significant momentum in recent years. Like Address Resolution Protocol (ARP) in IPv4, Neighbor Discovery Protocol (NDP) provides some functions like address resolution in IPv6. Besides functionality of NDP, it is vulnerable to some attacks. To mitigate these attacks, Internet Protocol Security (IPsec) was introduced, but it was not efficient due to its limitation. Therefore, SEND protocol is proposed to automatic protection of auto-configuration process. It is secure neighbor discovery and address resolution process. To defend against threats on NDP’s integrity and identity, Cryptographically Generated Address (CGA) and asymmetric cryptography are used by SEND. Besides advantages of SEND, its disadvantages like the computation process of CGA algorithm and sequentially of CGA generation algorithm are considerable. In this paper, we parallel this process between network resources in order to improve it. In addition, we compare the CGA generation time in self-computing and distributed-computing process. We focus on the impact of the malicious nodes on the CGA generation time in the network. According to the result, although malicious nodes participate in the generation process, CGA generation time is less than when it is computed in a one-way. By Trust Management System, detecting and insulating malicious nodes is easier.

Keywords: NDP, IPsec, SEND, CGA, Modifier, Malicious node, Self-Computing, Distributed-Computing.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1366

Authors: Zayed Freah Zeadat

Abstract:

The wellbeing of urban dwellers is strongly associated with the quality and quantity of green infrastructure. Nevertheless, urban green infrastructure is still lagging in many Arab cities, and Jordan is no exception. The capital city of Jordan, Amman, is becoming more urban dense with limited green spaces. The unplanned urban growth in Amman has caused several environmental problems such as urban heat islands, air pollution and lack of green spaces. This study aims to investigate the most suitable drivers to leverage the implementation of urban green infrastructure in Jordan through qualitative and quantitative analysis. The qualitative research includes an extensive literature review to discuss the most common drivers used internationally to promote urban green infrastructure implementation in the literature. The quantitative study employs a questionnaire survey to rank the suitability of each driver. Consultants, contractors and policymakers were invited to fill the research questionnaire according to their judgments and opinions. Relative Importance Index has been used to calculate the weighted average of all drivers and the Kruskal-Wallis test to check the degree of agreement among groups. This study finds that research participants agreed that indirect financial incentives (i.e., tax reductions, reduction in stormwater utility fee, reduction of interest rate, density bonus etc.) are the most effective incentive policy whilst granting sustainability certificate policy is the least effective driver to ensure widespread of UGI is elements in Jordan.

Keywords: sustainable development, urban green infrastructure, relative importance index, urban Jordan

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 551

Authors: V. Thulasi Bai, Srivatsa S. K.

Abstract:

Deaths from cardiovascular diseases have decreased substantially over the past two decades, largely as a result of advances in acute care and cardiac surgery. These developments have produced a growing population of patients who have survived a myocardial infarction. These patients need to be continuously monitored so that the initiation of treatment can be given within the crucial golden hour. The available conventional methods of monitoring mostly perform offline analysis and restrict the mobility of these patients within a hospital or room. Hence the aim of this paper is to design a Portable Cardiac Telemedicine System to aid the patients to regain their independence and return to an active work schedule, there by improving the psychological well being. The portable telemedicine system consists of a Wearable ECG Transmitter (WET) and a slightly modified mobile phone, which has an inbuilt ECG analyzer. The WET is placed on the body of the patient that continuously acquires the ECG signals from the high-risk cardiac patients who can move around anywhere. This WET transmits the ECG to the patient-s Bluetooth enabled mobile phone using blue tooth technology. The ECG analyzer inbuilt in the mobile phone continuously analyzes the heartbeats derived from the received ECG signals. In case of any panic condition, the mobile phone alerts the patients care taker by an SMS and initiates the transmission of a sample ECG signal to the doctor, via the mobile network.

Keywords: WET, ECG analyzer, Bluetooth, mobilecellular network, high risk cardiac patients.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2090

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 1241

Authors: Ahmet Kayabasi, Ali Akdagli

Abstract:

In this study, three robust predicting methods, namely artificial neural network (ANN), adaptive neuro fuzzy inference system (ANFIS) and support vector machine (SVM) were used for computing the resonant frequency of A-shaped compact microstrip antennas (ACMAs) operating at UHF band. Firstly, the resonant frequencies of 144 ACMAs with various dimensions and electrical parameters were simulated with the help of IE3D™ based on method of moment (MoM). The ANN, ANFIS and SVM models for computing the resonant frequency were then built by considering the simulation data. 124 simulated ACMAs were utilized for training and the remaining 20 ACMAs were used for testing the ANN, ANFIS and SVM models. The performance of the ANN, ANFIS and SVM models are compared in the training and test process. The average percentage errors (APE) regarding the computed resonant frequencies for training of the ANN, ANFIS and SVM were obtained as 0.457%, 0.399% and 0.600%, respectively. The constructed models were then tested and APE values as 0.601% for ANN, 0.744% for ANFIS and 0.623% for SVM were achieved. The results obtained here show that ANN, ANFIS and SVM methods can be successfully applied to compute the resonant frequency of ACMAs, since they are useful and versatile methods that yield accurate results.

Keywords: A-shaped compact microstrip antenna, Artificial Neural Network (ANN), adaptive Neuro-Fuzzy Inference System (ANFIS), Support Vector Machine (SVM).

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2209

Authors: Mochammad Dewo, Sumarsono Sudarto

Abstract:

The Exponential Triple Smoothing Algorithm approach nowadays, which is used to anticipate the export value of Indonesia's two major commodities, coal and palm oil, has a Mean Percentage Absolute Error (MAPE) value of 30-50%, which may be considered as a "reasonable" forecasting mistake. Forecasting errors of more than 30% shall have a domino effect on industrial output, as extra production adds to raw material, manufacturing and storage expenses. Whereas, reaching an "excellent" classification with an error value of less than 10% will provide new investors and exporters with confidence in the commercial development of related sectors. Industrial growth will bring out a positive impact on economic development. It can be applied for other commodities if the forecast error is less than 10%. The purpose of this project is to create a forecasting technique that can produce precise forecasting results with an error of less than 10%. This research analyzes forecasting methods such as ARIMA (Autoregressive Integrated Moving Average), ANN (Artificial Neural Network) and LSTM (Long-Short Term Memory). By providing a MAPE of 1%, this study reveals that ANN is the most successful strategy for forecasting coal and palm oil commodities in Indonesia.

Keywords: ANN, Artificial Neural Network, ARIMA, Autoregressive Integrated Moving Average, export value, forecast, LSTM, Long Short Term Memory.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 203

Authors: Endri Rama, Genci Capi, Shigenori Kawahara

Abstract:

The mobile robot ability to navigate autonomously in its environment is very important. Even though the advances in technology, robot self-localization and goal directed navigation in complex environments are still challenging tasks. In this article, we propose a novel method for robot navigation based on rat’s brain signals (Local Field Potentials). It has been well known that rats accurately and rapidly navigate in a complex space by localizing themselves in reference to the surrounding environmental cues. As the first step to incorporate the rat’s navigation strategy into the robot control, we analyzed the rats’ strategies while it navigates in a multiple Y-maze, and recorded Local Field Potentials (LFPs) simultaneously from three brain regions. Next, we processed the LFPs, and the extracted features were used as an input in the artificial neural network to predict the rat’s next location, especially in the decision-making moment, in Y-junctions. We developed an algorithm by which the robot learned to imitate the rat’s decision-making by mapping the rat’s brain signals into its own actions. Finally, the robot learned to integrate the internal states as well as external sensors in order to localize and navigate in the complex environment.

Keywords: Brain machine interface, decision-making, local field potentials, mobile robot, navigation, neural network, rat, signal processing.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1478

Authors: Pradeep Kumar, Tjoon Tow Teng, Shri Chand, Kailas L. Wasewar

Abstract:

The pulp and paper mill effluent is one of the high polluting effluent amongst the effluents obtained from polluting industries. All the available methods for treatment of pulp and paper mill effluent have certain drawbacks. The coagulation is one of the cheapest process for treatment of various organic effluents. Thus, the removal of chemical oxygen demand (COD) and colour of paper mill effluent is studied using coagulation process. The batch coagulation process was performed using various coagulants like: aluminium chloride, poly aluminium chloride and copper sulphate. The initial pH of the effluent (Coagulation pH) has tremendous effect on COD and colour removal. Poly aluminium chloride (PAC) as coagulant reduced COD to 84 % and 92 % of colour was removed at an optimum pH 5 and coagulant dose of 8 ml l-1. With aluminium chloride at an optimum pH = 4 and coagulant dose of 5 g l-1, 74 % COD and 86 % colour removal were observed. The results using copper sulphate as coagulant (a less commercial coagulant) were encouraging. At an optimum pH 6 and mass loading of 5 g l-1, 76 % COD reduction and 78 % colour reduction were obtained. It was also observed that after addition of coagulant, the pH of the effluent decreases. The decrease in pH was highest for AlCl3, which was followed by PAC and CuSO4. Significant amount of COD reductions was obtained by coagulation process. Since the coagulation process is the first stage for treatment of effluent and some of the coagulant cations usually remain in the treated effluents. Thus, cation like copper may be one of the good catalyst for second stage of treatment process like wet oxidation. The copper has been found to be good oxidation catalyst then iron and aluminum.

Keywords: Aluminium based coagulants, Coagulation, Copper, PAC, Pulp and paper mill effluent, Wastewater treatment

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 6984

Authors: Mohammad Yazdani, Mohammad Ali Bahmanyar, Hemmatollah Pirdashti, Mohammad Ali Esmaili

Abstract:

In order to study the effect of phosphate solubilization microorganisms (PSM) and plant growth promoting rhizobacteria (PGPR) on yield and yield components of corn Zea mays (L. cv. SC604) an experiment was conducted at research farm of Sari Agricultural Sciences and Natural Resources University, Iran during 2007. Experiment laid out as split plot based on randomized complete block design with three replications. Three levels of manures (consisted of 20 Mg.ha-1 farmyard manure, 15 Mg.ha-1 green manure and check or without any manures) as main plots and eight levels of biofertilizers (consisted of 1-NPK or conventional fertilizer application; 2-NPK+PSM+PGPR; 3 NP50%K+PSM+PGPR; 4- N50%PK+PSM +PGPR; 5-N50%P50%K+PSM+ PGPR; 6-PK+PGPR; 7- NK+PSM and 8-PSM+PGPR) as sub plots were treatments. Results showed that farmyard manure application increased row number, ear weight, grain number per ear, grain yield, biological yield and harvest index compared to check. Furthermore, using of PSM and PGPR in addition to conventional fertilizer applications (NPK) could improve ear weight, row number and grain number per row and ultimately increased grain yield in green manure and check plots. According to results in all fertilizer treatments application of PSM and PGPR together could reduce P application by 50% without any significant reduction of grain yield. However, this treatment could not compensate 50% reduction of N application.

Keywords: Biofertilizers, corn, PSM, PGPR, grain yield.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 6658

Authors: Bharathi P. T, P. Subashini

Abstract:

Ice cover County has a significant impact on rivers as it affects with the ice melting capacity which results in flooding, restrict navigation, modify the ecosystem and microclimate. River ices are made up of different ice types with varying ice thickness, so surveillance of river ice plays an important role. River ice types are captured using infrared imaging camera which captures the images even during the night times. In this paper the river ice infrared texture images are analysed using first-order statistical methods and secondorder statistical methods. The second order statistical methods considered are spatial gray level dependence method, gray level run length method and gray level difference method. The performance of the feature extraction methods are evaluated by using Probabilistic Neural Network classifier and it is found that the first-order statistical method and second-order statistical method yields low accuracy. So the features extracted from the first-order statistical method and second-order statistical method are combined and it is observed that the result of these combined features (First order statistical method + gray level run length method) provides higher accuracy when compared with the features from the first-order statistical method and second-order statistical method alone.

Keywords: Gray Level Difference Method, Gray Level Run Length Method, Kurtosis, Probabilistic Neural Network, Skewness, Spatial Gray Level Dependence Method.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2897

Authors: Vijayalakshmy Periaswamy S., Kevin McDaid

Abstract:

Testing accounts for the major percentage of technical contribution in the software development process. Typically, it consumes more than 50 percent of the total cost of developing a piece of software. The selection of software tests is a very important activity within this process to ensure the software reliability requirements are met. Generally tests are run to achieve maximum coverage of the software code and very little attention is given to the achieved reliability of the software. Using an existing methodology, this paper describes how to use Bayesian Belief Networks (BBNs) to select unit tests based on their contribution to the reliability of the module under consideration. In particular the work examines how the approach can enhance test-first development by assessing the quality of test suites resulting from this development methodology and providing insight into additional tests that can significantly reduce the achieved reliability. In this way the method can produce an optimal selection of inputs and the order in which the tests are executed to maximize the software reliability. To illustrate this approach, a belief network is constructed for a modern software system incorporating the expert opinion, expressed through probabilities of the relative quality of the elements of the software, and the potential effectiveness of the software tests. The steps involved in constructing the Bayesian Network are explained as is a method to allow for the test suite resulting from test-driven development.

Keywords: Software testing, Test Driven Development, Bayesian Belief Networks.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1880

Authors: Cameron Hamilton, Walter Potter, Gerrit Hoogenboom, Ronald McClendon, Will Hobbs

Abstract:

A model was constructed to predict the amount of solar radiation that will make contact with the surface of the earth in a given location an hour into the future. This project was supported by the Southern Company to determine at what specific times during a given day of the year solar panels could be relied upon to produce energy in sufficient quantities. Due to their ability as universal function approximators, an artificial neural network was used to estimate the nonlinear pattern of solar radiation, which utilized measurements of weather conditions collected at the Griffin, Georgia weather station as inputs. A number of network configurations and training strategies were utilized, though a multilayer perceptron with a variety of hidden nodes trained with the resilient propagation algorithm consistently yielded the most accurate predictions. In addition, a modeled direct normal irradiance field and adjacent weather station data were used to bolster prediction accuracy. In later trials, the solar radiation field was preprocessed with a discrete wavelet transform with the aim of removing noise from the measurements. The current model provides predictions of solar radiation with a mean square error of 0.0042, though ongoing efforts are being made to further improve the model’s accuracy.

Keywords: Artificial Neural Networks, Resilient Propagation, Solar Radiation, Time Series Forecasting.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2752

Authors: Fariba Shavakhi Zavareh, Hadi Fotoohabadi, Reza Sedaghati

Abstract:

Due to the continuous increment of the load demand, identification of weaker buses, improvement of voltage profile and power losses in the context of the voltage stability problems has become one of the major concerns for the larger, complex, interconnected power systems. The objective of this paper is to review the impact of Flexible AC Transmission System (FACTS) controller in Wind generators connected electrical network for maintaining voltage stability. Wind energy could be the growing renewable energy due to several advantages. The influence of wind generators on power quality is a significant issue; non uniform power production causes variations in system voltage and frequency. Therefore, wind farm requires high reactive power compensation; the advances in high power semiconducting devices have led to the development of FACTS. The FACTS devices such as for example SVC inject reactive power into the system which helps in maintaining a better voltage profile. The performance is evaluated on an IEEE 14 bus system, two wind generators are connected at low voltage buses to meet the increased load demand and SVC devices are integrated at the buses with wind generators to keep voltage stability. Power flows, nodal voltage magnitudes and angles of the power network are obtained by iterative solutions using MIPOWER.

Keywords: Voltage Profile, FACTS Device, SVC, Distributed Generation.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2656

Authors: Zaineb Ben Messaoud, Dorra Gargouri, Saida Zribi, Ahmed Ben Hamida

Abstract:

This paper presents a formant-tracking linear prediction (FTLP) model for speech processing in noise. The main focus of this work is the detection of formant trajectory based on Hidden Markov Models (HMM), for improved formant estimation in noise. The approach proposed in this paper provides a systematic framework for modelling and utilization of a time- sequence of peaks which satisfies continuity constraints on parameter; the within peaks are modelled by the LP parameters. The formant tracking LP model estimation is composed of three stages: (1) a pre-cleaning multi-band spectral subtraction stage to reduce the effect of residue noise on formants (2) estimation stage where an initial estimate of the LP model of speech for each frame is obtained (3) a formant classification using probability models of formants and Viterbi-decoders. The evaluation results for the estimation of the formant tracking LP model tested in Gaussian white noise background, demonstrate that the proposed combination of the initial noise reduction stage with formant tracking and LPC variable order analysis, results in a significant reduction in errors and distortions. The performance was evaluated with noisy natual vowels extracted from international french and English vocabulary speech signals at SNR value of 10dB. In each case, the estimated formants are compared to reference formants.

Keywords: Formants Estimation, HMM, Multi Band Spectral Subtraction, Variable order LPC coding, White Gauusien Noise.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1954

Authors: Yao-Hong Tsai

Abstract:

Due to the sensor technology, video surveillance has become the main way for security control in every big city in the world. Surveillance is usually used by governments for intelligence gathering, the prevention of crime, the protection of a process, person, group or object, or the investigation of crime. Many surveillance systems based on computer vision technology have been developed in recent years. Moving target tracking is the most common task for Unmanned Aerial Vehicle (UAV) to find and track objects of interest in mobile aerial surveillance for civilian applications. The paper is focused on vision-based collision avoidance for UAVs by recurrent neural networks. First, images from cameras on UAV were fused based on deep convolutional neural network. Then, a recurrent neural network was constructed to obtain high-level image features for object tracking and extracting low-level image features for noise reducing. The system distributed the calculation of the whole system to local and cloud platform to efficiently perform object detection, tracking and collision avoidance based on multiple UAVs. The experiments on several challenging datasets showed that the proposed algorithm outperforms the state-of-the-art methods.

Keywords: Unmanned aerial vehicle, object tracking, deep learning, collision avoidance.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 945