Search results for: back propagation neural network model

Authors: Zhiyan Meng, Dan Liu, Jintao Meng

Abstract:

Dependable encrypted traffic classification is crucial for improving cybersecurity and handling the growing amount of data. Large language models have shown that learning from large datasets can be effective, making pre-trained methods for encrypted traffic classification popular. However, attention-based pre-trained methods face two main issues: their large neural parameters are not suitable for low-computation environments like mobile devices and real-time applications, and they often overfit by getting stuck in local minima. To address these issues, we developed a lightweight transformer model, which reduces the computational parameters through lightweight vocabulary construction and Squeeze-and-Excitation Block. We use sharpness-aware optimization to avoid local minima during pre-training and capture temporal features with relative positional embeddings. Our approach keeps the model's classification accuracy high for downstream tasks. We conducted experiments on four datasets -USTC-TFC2016, VPN 2016, Tor 2016, and CICIOT 2022. Even with fewer than 18 million parameters, our method achieves classification results similar to methods with ten times as many parameters.

Keywords: sharpness-aware optimization, encrypted traffic classification, squeeze-and-excitation block, pretrained model

Procedia PDF Downloads 30

Authors: Go Bong Choi, Shin Je Lee, Sung Jin Yoo, Gibaek Lee, Jong Min Lee

Abstract:

In South Korea, it is difficult to obtain data for statistical pipe assessment. In this paper, to address these issues, we find that various statistical model presented before is how data mixed with noise and are whether apply in South Korea. Three major type of model is studied and if data is presented in the paper, we add noise to data, which affects how model response changes. Moreover, we generate data from model in paper and analyse effect of noise. From this we can find robustness and applicability in Korea of each model.

Keywords: proportional hazard model, survival model, water main deterioration, ecological sciences

Procedia PDF Downloads 743

Authors: Zouhair Haddi, Stephane Delliaux, Jean-Francois Pons, Ismail Kechaf, Jean-Claude De Haro, Mustapha Ouladsine

Abstract:

Atrial fibrillation (AF) has been considered as the most common cardiac arrhythmia, and a major public health burden associated with significant morbidity and mortality. Nowadays, telemedical approaches targeting cardiac outpatients situate AF among the most challenged medical issues. The automatic, early, and fast AF detection is still a major concern for the healthcare professional. Several algorithms based on univariate analysis have been developed to detect atrial fibrillation. However, the published results do not show satisfactory classification accuracy. This work was aimed at resolving this shortcoming by proposing multivariate data analysis methods for automatic AF detection. Four publicly-accessible sets of clinical data (AF Termination Challenge Database, MIT-BIH AF, Normal Sinus Rhythm RR Interval Database, and MIT-BIH Normal Sinus Rhythm Databases) were used for assessment. All time series were segmented in 1 min RR intervals window and then four specific features were calculated. Two pattern recognition methods, i.e., Principal Component Analysis (PCA) and Learning Vector Quantization (LVQ) neural network were used to develop classification models. PCA, as a feature reduction method, was employed to find important features to discriminate between AF and Normal Sinus Rhythm. Despite its very simple structure, the results show that the LVQ model performs better on the analyzed databases than do existing algorithms, with high sensitivity and specificity (99.19% and 99.39%, respectively). The proposed AF detection holds several interesting properties, and can be implemented with just a few arithmetical operations which make it a suitable choice for telecare applications.

Keywords: atrial fibrillation, multivariate data analysis, automatic detection, telemedicine

Procedia PDF Downloads 267

Authors: Aaron Aboshio

Abstract:

Numerical study of the static response of homogeneous clay stratum considering a wide range of cohesion and subject to foundation loads is presented. The linear elastic–perfectly plastic constitutive relation with the von Mises yield criterion were utilised to develop a numerically cost effective finite element model for the soil while imposing a rigid body constrain to the foundation footing. From the analyses carried out, estimate of the bearing capacity factor, Nc as well as the ultimate load-carrying capacities of these soils, effect of cohesion on foundation settlements, stress fields and failure propagation were obtained. These are consistent with other findings in the literature and hence can be a useful guide in design of safe foundations in clay soils for buildings and other structure.

Keywords: bearing capacity factors, finite element method, safe bearing pressure, structure-soil interaction

Procedia PDF Downloads 302

Authors: Jurijs Salijevs, Katrina Bolocko

Abstract:

The integration of artificial intelligence and neural networks has significantly changed full-body imaging and high-resolution 3D mapping systems, and this paper reviews research in these areas. With an emphasis on their use in the early identification of melanoma and other disorders, the goal is to give a wide perspective on the current status and potential future of these medical imaging technologies. Authors also examine methodologies such as machine learning and deep learning, seeking to identify efficient procedures that enhance diagnostic capabilities through the analysis of 3D body scans. This work aims to encourage further research and technological development to harness the full potential of AI in disease diagnosis.

Keywords: artificial intelligence, neural networks, 3D scan, body scan, 3D mapping system, healthcare

Procedia PDF Downloads 103

Authors: Shiri Elisheva Decktor, Sharon Hornstein

Abstract:

Highway merges and unmarked junctions are key components in any urban road network, which can act as bottlenecks and create traffic disruption. Inefficient highway merges may trigger traffic instabilities such as stop-and-go waves, pose safety conditions and lead to longer journey times. These phenomena occur spontaneously if the average vehicle density exceeds a certain critical value. This study focuses on modeling the traffic using a microscopic traffic flow model. A hybrid traffic model, which combines human-driven and controlled vehicles is assumed. The controlled vehicles obey different driving policies when approaching the merge, or in the vicinity of other vehicles. We developed a co-simulation model in SUMO (Simulation of Urban Mobility), in which the human-driven cars are modeled using the IDM model, and the controlled cars are modeled using a dedicated controller. The scenario chosen for this study is a closed track with one merge and one exit, which could be later implemented using a scaled infrastructure on our lab setup. This will enable us to benchmark the results of this study obtained in simulation, to comparable results in similar conditions in the lab. The metrics chosen for the comparison of the performance of our algorithm on the overall traffic conditions include the average speed, wait time near the merge, and throughput after the merge, measured under different travel demand conditions (low, medium, and heavy traffic).

Keywords: highway merges, traffic modeling, SUMO, driving policy

Procedia PDF Downloads 106

Authors: Amjad Khan

Abstract:

The security of Vehicular ad hoc networking is of great importance as it involves serious life threats. Thus to provide secure communication amongst Vehicles on road, the conventional security system is not enough. It is necessary to prevent the network resources from wastage and give them protection against malicious nodes so that to ensure the data bandwidth availability to the legitimate nodes of the network. This work is related to provide a non conventional security system by introducing some constraints to minimize the DoS (Denial of services) especially data and bandwidth. The data packets received by a node in the network will pass through a number of tests and if any of the test fails, the node will drop those data packets and will not forward it anymore. Also if a node claims to be the nearest node for forwarding emergency messages then the sender can effectively identify the true or false status of the claim by using these constraints. Consequently the DoS(Denial of Services) attack is minimized by the instant availability of data without wasting the network resources.

Keywords: black hole attack, grey hole attack, intransient traffic tempering, networking

Procedia PDF Downloads 284

Authors: Ashwani Kumar

Abstract:

Use of NIR light for imaging the biological tissue and to quantify its optical properties is a good choice over other invasive methods. Optical tomography involves two steps. One is the forward problem and the other is the reconstruction problem. The forward problem consists of finding the measurements of transmitted light through the tissue from source to detector, given the spatial distribution of absorption and scattering properties. The second step is the reconstruction problem. In X-ray tomography, there is standard method for reconstruction called filtered back projection method or the algebraic reconstruction methods. But this method cannot be applied as such, in optical tomography due to highly scattering nature of biological tissue. A hybrid algorithm for reconstruction has been implemented in this work which takes into account the highly scattered path taken by photons while back projecting the forward data obtained during Monte Carlo simulation. The reconstructed image suffers from blurring due to point spread function.

Keywords: NIR light, tissue, blurring, Monte Carlo simulation

Procedia PDF Downloads 494

Authors: Arnab Mondal, Sanjeev Kumar Sharma, Ranjit Kumar Upadhyay

Abstract:

We study the spatiotemporal dynamics of a neuronal cable. The processes of one- dimensional (1D) and 2D diffusion are considered for a single variable, which is the membrane voltage, i.e., membrane voltage diffusively interacts for spatiotemporal pattern formalism. The recovery and other variables interact through the membrane voltage. A 3D Leech-Heart (LH) model is introduced to investigate the nonlinear responses of an excitable neuronal cable. The deterministic LH model shows different types of firing properties. We explore the parameter space of the uncoupled LH model and based on the bifurcation diagram, considering v_k2_ashift as a bifurcation parameter, we analyze the 1D diffusion dynamics in three regimes: bursting, regular spiking, and a quiescent state. Depending on parameters, it is shown that the diffusive system may generate regular and irregular bursting or spiking behavior. Further, it is explored a 2D diffusion acting on the membrane voltage, where different types of patterns can be observed. The results show that the LH neurons with different firing characteristics depending on the control parameters participate in a collective behavior of an information processing system that depends on the overall network.

Keywords: bifurcation, pattern formation, spatio-temporal dynamics, stability analysis

Procedia PDF Downloads 222

Authors: Brandon Foggo, Nanpeng Yu

Abstract:

Power distribution circuits undergo frequent network topology changes that are often left undocumented. As a result, the documentation of a circuit’s connectivity becomes inaccurate with time. The lack of reliable circuit connectivity information is one of the biggest obstacles to model, monitor, and control modern distribution systems. To enhance the reliability and efficiency of electric power distribution systems, the circuit’s connectivity information must be updated periodically. This paper focuses on one critical component of a distribution circuit’s topology - the secondary transformer to phase association. This topology component describes the set of phase lines that feed power to a given secondary transformer (and therefore a given group of power consumers). Finding the documentation of this component is call Phase Identification, and is typically performed with physical measurements. These measurements can take time lengths on the order of several months, but with supervised learning, the time length can be reduced significantly. This paper compares several such methods applied to Phase Identification for a large range of real distribution circuits, describes a method of training data selection, describes preprocessing steps unique to the Phase Identification problem, and ultimately describes a method which obtains high accuracy (> 96% in most cases, > 92% in the worst case) using only 5% of the measurements typically used for Phase Identification.

Keywords: distribution network, machine learning, network topology, phase identification, smart grid

Procedia PDF Downloads 299

Authors: O. I. Nkwonta, B. Dzwairo, J. Adeyemo, A. Jaiyola, N. Sawyerr, F. Otieno

Abstract:

The effective management of water resources is of great importance to ensure the supply of water resources to support changing water requirements over a selected planning horizon and in a sustainable and cost-effective way. Essentially, the purpose of the water resources planning process is to balance the available water resources in a system with the water requirements and losses to which the system is subjected. In such situations, Water resources yield and planning model can be used to solve those difficulties. It has an advantage over other models by managing model runs, developing a representative system network, modelling incremental sub-catchments, creating a variety of standard system features, special modelling features, and run result output options.

Keywords: complex, water resources, planning, cost effective and management

Procedia PDF Downloads 573

Authors: Chung-Yuan Huang, Yu-Hsiang Fu, Chuen-Tsai Sun

Abstract:

Community detection represents an important data-mining tool for analyzing and understanding real-world complex network structures and functions. We believe that at least four criteria determine the appropriateness of a community detection algorithm: (a) it produces useable normalized mutual information (NMI) and modularity results for social networks, (b) it overcomes resolution limitation problems associated with synthetic networks, (c) it produces good NMI results and performance efficiency for Lancichinetti-Fortunato-Radicchi (LFR) benchmark networks, and (d) it produces good modularity and performance efficiency for large-scale real-world complex networks. To our knowledge, no existing community detection algorithm meets all four criteria. In this paper, we describe a simple hierarchical arc-merging (HAM) algorithm that uses network topologies and rule-based arc-merging strategies to identify community structures that satisfy the criteria. We used five well-studied social network datasets and eight sets of LFR benchmark networks to validate the ground-truth community correctness of HAM, eight large-scale real-world complex networks to measure its performance efficiency, and two synthetic networks to determine its susceptibility to resolution limitation problems. Our results indicate that the proposed HAM algorithm is capable of providing satisfactory performance efficiency and that HAM-identified communities were close to ground-truth communities in social and LFR benchmark networks while overcoming resolution limitation problems.

Keywords: complex network, social network, community detection, network hierarchy

Procedia PDF Downloads 227

Authors: Xi Liu, Fan Yuan

Abstract:

In the evolving landscape of physics-based machine learning (PBML), particularly within fluid dynamics and its applications in electromechanical engineering, robot vision, and robot learning, achieving precision and alignment with researchers' specific needs presents a formidable challenge. In response, this work proposes a methodology that integrates neural transformation with a modified smoothed particle hydrodynamics model for generating transformed 3D fluid simulations. This approach is useful for nanoscale science, where the unique and complex behaviors of viscoelastic medium demand accurate neurally-transformed simulations for materials understanding and manipulation. In electromechanical engineering, the method enhances the design and functionality of fluid-operated systems, particularly microfluidic devices, contributing to advancements in nanomaterial design, drug delivery systems, and more. The proposed approach also aligns with the principles of PBML, offering advantages such as multi-fluid stylization and consistent particle attribute transfer. This capability is valuable in various fields where the interaction of multiple fluid components is significant. Moreover, the application of neurally-transformed hydrodynamical models extends to manufacturing processes, such as the production of microelectromechanical systems, enhancing efficiency and cost-effectiveness. The system's ability to perform neural transfer on 3D fluid scenes using a deep learning algorithm alongside physical models further adds a layer of flexibility, allowing researchers to tailor simulations to specific needs across scientific and engineering disciplines.

Keywords: physics-based machine learning, robot vision, robot learning, hydrodynamics

Procedia PDF Downloads 66

Authors: Aydin Teymourifar, Gurkan Ozturk, Ozan Bahadir

Abstract:

NSGA-II is one of the most well-known and most widely used evolutionary algorithms. In addition to its new versions, such as NSGA-III, there are several modified types of this algorithm in the literature. In this paper, a hybrid NSGA-II algorithm has been suggested for solving the multi-objective flexible job shop scheduling problem. For a better search, new neighborhood-based crossover and mutation operators are defined. To create new generations, the neighbors of the selected individuals by the tournament selection are constructed. Also, at the end of each iteration, before sorting, neighbors of a certain number of good solutions are derived, except for solutions protected by elitism. The neighbors are generated using a constraint-based neural network that uses various constructs. The non-dominated sorting and crowding distance operators are same as the classic NSGA-II. A comparison based on some multi-objective benchmarks from the literature shows the efficiency of the algorithm.

Keywords: flexible job shop scheduling problem, multi-objective optimization, NSGA-II algorithm, neighborhood structures

Procedia PDF Downloads 229

Authors: Panagiotis Svarnas, Polykarpos Papadopoulos

Abstract:

Atmospheric-pressure cold plasmas continue to gain increasing interest for various applications due to their unique properties, like cost-efficient production, high chemical reactivity, low gas temperature, adaptability, etc. Numerous designs have been proposed for these plasmas production in terms of electrode configuration, driving voltage waveform and working gas(es). However, in order to exploit most of the advantages of these systems, the majority of the designs are based on dielectric-barrier discharges (DBDs) either in filamentary or glow regimes. A special category of the DBD-based atmospheric-pressure cold plasmas refers to the so-called plasma jets, where a carrier noble gas is guided by the dielectric barrier (usually a hollow cylinder) and left to flow up to the atmospheric air where a complicated hydrodynamic interplay takes place. Although it is now well established that these plasmas are generated due to ionizing waves reminding in many ways streamer propagation, they exhibit discrete characteristics which are better mirrored on the terms 'guided streamers' or 'plasma bullets'. These 'bullets' travel with supersonic velocities both inside the dielectric barrier and the channel formed by the noble gas during its penetration into the air. The present work is devoted to the interpretation of the electro-hydrodynamic effects that take place downstream of the dielectric barrier opening, i.e., in the noble gas-air mixing area where plasma bullet propagate under the influence of local electric fields in regions of variable noble gas concentration. Herein, we focus on the role of the local space charge and the residual ionic charge left behind after the bullet propagation in the gas flow field modification. The study communicates both experimental and numerical results, coupled in a comprehensive manner. The plasma bullets are here produced by a custom device having a quartz tube as a dielectric barrier and two external ring-type electrodes driven by sinusoidal high voltage at 10 kHz. Helium gas is fed to the tube and schlieren photography is employed for mapping the flow field downstream of the tube orifice. Mixture mass conservation equation, momentum conservation equation, energy conservation equation in terms of temperature and helium transfer equation are simultaneously solved, leading to the physical mechanisms that govern the experimental results. Namely, we deal with electro-hydrodynamic effects mainly due to momentum transfer from atomic ions to neutrals. The atomic ions are left behind as residual charge after the bullet propagation and gain energy from the locally created electric field. The electro-hydrodynamic force is eventually evaluated.

Keywords: atmospheric-pressure plasmas, dielectric-barrier discharges, schlieren photography, electro-hydrodynamic force

Procedia PDF Downloads 139

Authors: Babesse Saad, Ameddah Djemeleddine

Abstract:

In this paper, a learning algorithm using neuronal networks to improve the roll stability and prevent the rollover in a single unit heavy vehicle is proposed. First, LQR control to keep balanced normalized rollovers, between front and rear axles, below the unity, then a data collected from this controller is used as a training basis of a neuronal regulator. The ANN controller is thereafter applied for the nonlinear side force model, and gives satisfactory results than the LQR one.

Keywords: rollover, single unit heavy vehicle, neural networks, nonlinear side force

Procedia PDF Downloads 474

Authors: Farhan Iqbal

Abstract:

Ergonomics has its implication in classroom. Present study aimed at finding out the comfort level of students at university level due to classroom furniture which may affect students learning. Two public and one private institution was targeted. Purposive sampling was done. Four hundred and seventy five students volunteered to reply to a questionnaire. Different furniture were measured and descriptively compared with ISO 5970 standard. Overall discomfort was found to be statistically significant as compared to comfort. Comfort and discomfort were found to be negatively correlated. Gender did not differ on upper body discomfort, though, the median score found men to be more comfortable at upper body. GPA was found to be independent of comfort level. Most afflicted areas were neck, shoulder, upper back, lower back and pelvic. The present study will be helpful for all educational institutions of Pakistan. Future studies may be carried out with structural and functional anthropometric data of students for redesigning of the classroom furniture.

Keywords: anthropometry, classroom furniture, comfort, discomfort, learning

Procedia PDF Downloads 311

Authors: Louise Moles, Steve Riley, Sarah D. Lichenstein, Marzieh Babaeianjelodar, Robert Kohler, Annie Cheng, Corey Horien Abigail Greene, Wenjing Luo, Jonathan Ahern, Bohan Xu, Yize Zhao, Chun Chieh Fan, R. Todd Constable, Sarah W. Yip

Abstract:

Background: Risks for depression are myriad and include both genetic and brain-based factors. However, relationships between these systems are poorly understood, limiting understanding of disease etiology, particularly at the developmental level. Methods: We use a data-driven machine learning approach connectome-based predictive modeling (CPM) to identify functional connectivity signatures associated with polygenic risk scores for depression (DEP-PRS) among youth from the Adolescent Brain and Cognitive Development (ABCD) study across diverse brain states, i.e., during resting state, during affective working memory, during response inhibition, during reward processing. Results: Using 10-fold cross-validation with 100 iterations and permutation testing, CPM identified connectivity signatures of DEP-PRS across all examined brain states (rho’s=0.20-0.27, p’s<.001). Across brain states, DEP-PRS was positively predicted by increased connectivity between frontoparietal and salience networks, increased motor-sensory network connectivity, decreased salience to subcortical connectivity, and decreased subcortical to motor-sensory connectivity. Subsampling analyses demonstrated that model accuracies were robust across random subsamples of N’s=1,000, N’s=500, and N’s=250 but became unstable at N’s=100. Conclusions: These data, for the first time, identify neural networks of polygenic depression risk in a large sample of youth before the onset of significant clinical impairment. Identified networks may be considered potential treatment targets or vulnerability markers for depression risk.

Keywords: genetics, functional connectivity, pre-adolescents, depression

Procedia PDF Downloads 58

Authors: Surapol Raadnui

Abstract:

This particular work describes an experimental investigation on gear wear in which wear and pitting were intentionally allowed to occur, namely, moisture corrosion pitting, acid-induced corrosion pitting, hard contaminant-related pitting and mechanical induced wear. A back to back spur gear test rig and a grease lubricated worm gear rig were used. The tests samples of wear debris were collected and assessed through the utilization of an optical microscope in order to correlate and compare the debris morphology to pitting and wear degradation of the worn gears. In addition, weight loss from all test gear pairs were assessed with utilization of statistical design of experiment. It can be deduced that wear debris characteristics from both cases exhibited a direct relationship with different pitting and wear modes. Thus, it should be possible to detect and diagnose gear pitting and wear utilization of worn surfaces, generated wear debris and quantitative measurement such as weight loss.

Keywords: predictive maintenance, worm gear, spur gear, wear debris analysis, problem diagnostic

Procedia PDF Downloads 153

Authors: Xiping Wang，Shuran Yao, Liqin Dai

Abstract:

Many studies report the winter is getting warmer and the minimum air temperature is obviously rising as the important climate warming evidences. The exacerbated air temperature fluctuation tending to bring more severe weather variation is another important consequence of recent climate change which induced more disasters to crop growth in quite a certain regions. Hebei Province is an important winter wheat growing province in North of China that recently endures more winter freezing injury influencing the local winter wheat crop management. A winter wheat freezing injury assessment Bayesian Network framework was established for the objectives of estimating, assessing and predicting winter wheat freezing disasters in Hebei Province. In this framework, the freezing disasters was classified as three severity degrees (SI) among all the three types of freezing, i.e., freezing caused by severe cold in anytime in the winter, long extremely cold duration in the winter and freeze-after-thaw in early season after winter. The factors influencing winter wheat freezing SI include time of freezing occurrence, growth status of seedlings, soil moisture, winter wheat variety, the longitude of target region and, the most variable climate factors. The climate factors included in this framework are daily mean and range of air temperature, extreme minimum temperature and number of days during a severe cold weather process, the number of days with the temperature lower than the critical temperature values, accumulated negative temperature in a potential freezing event. The Bayesian Network model was evaluated using actual weather data and crop records at selected sites in Hebei Province using real data. With the multi-stage influences from the various factors, the forecast and assessment of the event-based target variables, freezing injury occurrence and its damage to winter wheat production, were shown better scoped by Bayesian Network model.

Keywords: bayesian networks, climatic change, freezing Injury, winter wheat

Procedia PDF Downloads 408

Authors: S. Vignesh, K. S. Rangasamy

Abstract:

The paper deals with the device capable of detecting and disabling digital cameras. The system locates the camera and then neutralizes it. Every digital camera has an image sensor known as a CCD, which is retro-reflective and sends light back directly to its original source at the same angle. The device shines infrared LED light, which is invisible to the human eye, at a distance of about 20 feet. It then collects video of these reflections with a camcorder. Then the video of the reflections is transferred to a computer connected to the device, where it is sent through image processing algorithms that pick out infrared light bouncing back. Once the camera is detected, the device would project an invisible infrared laser into the camera's lens, thereby overexposing the photo and rendering it useless. Low levels of infrared laser neutralize digital cameras but are neither a health danger to humans nor a physical damage to cameras. We also discuss the simplified design of the above device that can used in theatres to prevent piracy. The domains being covered here are optics and image processing.

Keywords: CCD, optics, image processing, D3CIP

Procedia PDF Downloads 357

Authors: Nirmal Kumar Verma, Pallavi Jha

Abstract:

Terahertz (THz) radiation generation by propagation of two-color laser pulses in plasma is an active area of research due to its potential applications in various areas, including security screening, material characterization and spectroscopic techniques. Due to non ionizing nature and the ability to penetrate several millimeters, THz radiation is suitable for diagnosis of cancerous cells. Traditional THz emitters like optically active crystals when irradiated with high power laser radiation, are subject to material breakdown and hence low conversion efficiencies. This problem is not encountered in laser - plasma based THz radiation sources. The present paper is devoted to the simulation study of the enhanced THz radiation generation by propagation of two-color, linearly polarized laser pulses through magnetized plasma. The two laser pulses orthogonally polarized are co-propagating along the same direction. The direction of the external magnetic field is such that one of the two laser pulses propagates in the ordinary mode, while the other pulse propagates in the extraordinary mode through homogeneous plasma. A transverse electromagnetic wave with frequency in the THz range is generated due to the presence of the static magnetic field. It is observed that larger amplitude terahertz can be generated by mixing of ordinary and extraordinary modes of two-color laser pulses as compared with a single laser pulse propagating in the extraordinary mode.

Keywords: two-color laser pulses, terahertz radiation, magnetized plasma, ordinary and extraordinary mode

Procedia PDF Downloads 301

Authors: Leah Wright, Trevor Townsend

Abstract:

Mode choice models are crucial instruments in the analysis of travel behaviour. These models show the relationship between an individual’s choice of transportation mode for a given O-D pair and the individual’s socioeconomic characteristics such as household size and income level, age and/or gender, and the features of the transportation system. The most popular functional forms of these models are based on Utility-Based Choice Theory, which addresses the uncertainty in the decision-making process with the use of an error term. However, with the development of artificial intelligence, many researchers have started to take a different approach to travel demand modelling. In recent times, researchers have looked at using neural networks, fuzzy logic and rough set theory to develop improved mode choice formulas. The concept of cloud theory has recently been introduced to model decision-making under uncertainty. Unlike the previously mentioned theories, cloud theory recognises a relationship between randomness and fuzziness, two of the most common types of uncertainty. This research aims to investigate the use of cloud theory in mode choice models. This paper highlights the conceptual framework of the mode choice model using cloud theory. Merging decision-making under uncertainty and mode choice models is state of the art. The cloud theory model is expected to address the issues and concerns with the nested logit and improve the design of mode choice models and their use in travel demand.

Keywords: Cloud theory, decision-making, mode choice models, travel behaviour, uncertainty

Procedia PDF Downloads 388

Authors: Mamouni Mohammed Dhiya Eddine

Abstract:

This work aims at studying the exploitation of high-speed networks of clusters for distributed storage. Parallel applications running on clusters require both high-performance communications between nodes and efficient access to the storage system. Many studies on network technologies led to the design of dedicated architectures for clusters with very fast communications between computing nodes. Efficient distributed storage in clusters has been essentially developed by adding parallelization mechanisms so that the server(s) may sustain an increased workload. In this work, we propose to improve the performance of distributed storage systems in clusters by efficiently using the underlying high-performance network to access distant storage systems. The main question we are addressing is: do high-speed networks of clusters fit the requirements of a transparent, efficient and high-performance access to remote storage? We show that storage requirements are very different from those of parallel computation. High-speed networks of clusters were designed to optimize communications between different nodes of a parallel application. We study their utilization in a very different context, storage in clusters, where client-server models are generally used to access remote storage (for instance NFS, PVFS or LUSTRE). Our experimental study based on the usage of the GM programming interface of MYRINET high-speed networks for distributed storage raised several interesting problems. Firstly, the specific memory utilization in the storage access system layers does not easily fit the traditional memory model of high-speed networks. Secondly, client-server models that are used for distributed storage have specific requirements on message control and event processing, which are not handled by existing interfaces. We propose different solutions to solve communication control problems at the filesystem level. We show that a modification of the network programming interface is required. Data transfer issues need an adaptation of the operating system. We detail several propositions for network programming interfaces which make their utilization easier in the context of distributed storage. The integration of a flexible processing of data transfer in the new programming interface MYRINET/MX is finally presented. Performance evaluations show that its usage in the context of both storage and other types of applications is easy and efficient.

Keywords: distributed storage, remote file access, cluster, high-speed network, MYRINET, zero-copy, memory registration, communication control, event notification, application programming interface

Procedia PDF Downloads 219

Authors: Wennan Long, Yuhao Nie, Yunan Li, Adam Brandt

Abstract:

To fight against climate change, California government issued the Senate Bill No. 100 (SB-100) in 2018 September, which aims at achieving a target of 100% renewable electricity by the end of 2045. A capacity expansion problem is solved in this case study using a binary quadratic programming model. The optimal locations and capacities of the potential renewable power plants (i.e., solar, wind, biomass, geothermal and hydropower), the phase-out schedule of existing fossil-based (nature gas) power plants and the transmission of electricity across the entire network are determined with the minimal total annualized cost measured by net present value (NPV). The results show that the renewable electricity contribution could increase to 85.9% by 2030 and reach 100% by 2035. Fossil-based power plants will be totally phased out around 2035 and solar and wind will finally become the most dominant renewable energy resource in California electricity mix.

Keywords: 100% renewable electricity, California, capacity expansion, mixed integer non-linear programming

Procedia PDF Downloads 171

Authors: Chen Jun-Hong, He Pu, Tao Wen-Quan

Abstract:

Proton exchange membrane fuel cell (PEMFC) is the most promising future energy source owing to its low operating temperature, high energy efficiency, high power density, and environmental friendliness. In this paper, a comprehensive PEMFC system control-oriented model is developed in the Matlab/Simulink environment, which includes the hydrogen supply subsystem, air supply subsystem, and thermal management subsystem. Besides, Improved Artificial Bee Colony (IABC) is used in the parameter identification of PEMFC semi-empirical equations, making the maximum relative error between simulation data and the experimental data less than 0.4%. Operation temperature is essential for PEMFC, both high and low temperatures are disadvantageous. In the thermal management subsystem, water pump and fan are both controlled with the PID controller to maintain the appreciate operation temperature of PEMFC for the requirements of safe and efficient operation. To improve the control effect further, fuzzy control is introduced to optimize the PID controller of the pump, and the Radial Basis Function (RBF) neural network is introduced to optimize the PID controller of the fan. The results demonstrate that Fuzzy-PID and RBF-PID can achieve a better control effect with 22.66% decrease in Integral Absolute Error Criterion (IAE) of T_st (Temperature of PEMFC) and 77.56% decrease in IAE of T_in (Temperature of inlet cooling water) compared with traditional PID. In the end, a novel thermal management structure is proposed, which uses the cooling air passing through the main radiator to continue cooling the secondary radiator. In this thermal management structure, the parasitic power dissipation can be reduced by 69.94%, and the control effect can be improved with a 52.88% decrease in IAE of T_in under the same controller.

Keywords: PEMFC system, parameter identification, temperature control, Fuzzy-PID, RBF-PID, parasitic power

Procedia PDF Downloads 85

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: Natascha C. D'Amico, Enzo Grossi, Giovanni Valbusa, Ala Malasevschi, Gianpiero Cardone, Sergio Papa

Abstract:

Purpose: To characterize, through a radiomic approach, the nature of areas classified PI-RADS (Prostate Imaging Reporting and Data System) 3/5, recognized in multiparametric prostate magnetic resonance with T2-weighted (T2w), diffusion and perfusion sequences with paramagnetic contrast. Methods and Materials: 24 cases undergoing multiparametric prostate MR and biopsy were admitted to this pilot study. Clinical outcome of the PI-RADS 3/5 was found through biopsy, finding 8 malignant tumours. The analysed images were acquired with a Philips achieva 1.5T machine with a CE- T2-weighted sequence in the axial plane. Semi-automatic tumour segmentation was carried out on MR images using 3DSlicer image analysis software. 45 shape-based, intensity-based and texture-based features were extracted and represented the input for preprocessing. An evolutionary algorithm (a TWIST system based on KNN algorithm) was used to subdivide the dataset into training and testing set and select features yielding the maximal amount of information. After this pre-processing 20 input variables were selected and different machine learning systems were used to develop a predictive model based on a training testing crossover procedure. Results: The best machine learning system (three-layers feed-forward neural network) obtained a global accuracy of 90% ( 80 % sensitivity and 100% specificity ) with a ROC of 0.82. Conclusion: Machine learning systems coupled with radiomics show a promising potential in distinguishing benign from malign tumours in PI-RADS 3/5 areas.

Keywords: machine learning, MR prostate, PI-Rads 3, radiomics

Procedia PDF Downloads 188

Authors: Hailyie Tekleselase

Abstract:

Deep learning is increasingly used as a building block of security systems. However, neural networks are hard to interpret and typically solid to the practitioner. This paper presents a detail survey of computing methods in cyber security, and analyzes the prospects of enhancing the cyber security capabilities by suggests that of accelerating the intelligence of the security systems. There are many AI-based applications used in industrial scenarios such as Internet of Things (IoT), smart grids, and edge computing. Machine learning technologies require a training process which introduces the protection problems in the training data and algorithms. We present machine learning techniques currently applied to the detection of intrusion, malware, and spam. Our conclusions are based on an extensive review of the literature as well as on experiments performed on real enterprise systems and network traffic. We conclude that problems can be solved successfully only when methods of artificial intelligence are being used besides human experts or operators.

Keywords: artificial intelligence, machine learning, deep learning, cyber security, big data

Procedia PDF Downloads 126

Authors: Narava Manose

Abstract:

Distillation is a very old separation technology for separating liquid mixtures that can be traced back to the chemists in Alexandria in the first century A. D. Today distillation is the most important industrial separation technology. By the eleventh century, distillation was being used in Italy to produce alcoholic beverages. At that time, distillation was probably a batch process based on the use of just a single stage, the boiler. The word distillation is derived from the Latin word destillare, which means dripping or trickling down. By at least the sixteenth century, it was known that the extent of separation could be improved by providing multiple vapor-liquid contacts (stages) in a so called Rectifactorium. The term rectification is derived from the Latin words rectefacere, meaning to improve. Modern distillation derives its ability to produce almost pure products from the use of multi-stage contacting. Throughout the twentieth century, multistage distillation was by far the most widely used industrial method for separating liquid mixtures of chemical components.The basic principle behind this technique relies on the different boiling temperatures for the various components of the mixture, allowing the separation between the vapor from the most volatile component and the liquid of other(s) component(s). •Developed a simple non-linear model of a binary distillation column using Skogestad equations in Simulink. •We have computed the steady-state operating point around which to base our analysis and controller design. However, the model contains two integrators because the condenser and reboiler levels are not controlled. One particular way of stabilizing the column is the LV-configuration where we use D to control M_D, and B to control M_B; such a model is given in cola_lv.m where we have used two P-controllers with gains equal to 10.

Keywords: modelling, distillation column, control, binary distillation

Procedia PDF Downloads 277