Search results for: Kolmogorov Arnold networks

Authors: Sofia Dokuka, Liz Lockhart, Alex Furman

Abstract:

Organizational hierarchy, traditionally composed of individual contributors, middle management, and executives, is enhanced by the understanding of informal social roles. These roles, identified with organizational network analysis (ONA), might have an important effect on organizational functioning. In this paper, we identify three social roles – influencers, bridges, and soloists, and provide empirical analysis based on real-world organizational networks. Influencers are employees with broad networks and whose contacts also have rich networks. Influence is calculated using PageRank, initially proposed for measuring website importance, but now applied in various network settings, including social networks. Influencers, having high PageRank, become key players in shaping opinions and behaviors within an organization. Bridges serve as links between loosely connected groups within the organization. Bridges are identified using betweenness and Burt’s constraint. Betweenness quantifies a node's control over information flows by evaluating its role in the control over the shortest paths within the network. Burt's constraint measures the extent of interconnection among an individual's contacts. A high constraint value suggests fewer structural holes and lesser control over information flows, whereas a low value suggests the contrary. Soloists are individuals with fewer than 5 stable social contacts, potentially facing challenges due to reduced social interaction and hypothetical lack of feedback and communication. We considered social roles in the analysis of real-world organizations (N=1,060). Based on data from digital traces (Slack, corporate email and calendar) we reconstructed an organizational communication network and identified influencers, bridges and soloists. We also collected employee engagement data through an online survey. Among the top-5% of influencers, 10% are members of the Executive Team. 56% of the Executive Team members are part of the top influencers group. The same proportion of top influencers (10%) is individual contributors, accounting for just 0.6% of all individual contributors in the company. The majority of influencers (80%) are at the middle management level. Out of all middle managers, 19% hold the role of influencers. However, individual contributors represent a small proportion of influencers, and having information about these individuals who hold influential roles can be crucial for management in identifying high-potential talents. Among the bridges, 4% are members of the Executive Team, 16% are individual contributors, and 80% are middle management. Predominantly middle management acts as a bridge. Bridge positions of some members of the executive team might indicate potential micromanagement on the leader's part. Recognizing the individuals serving as bridges in an organization uncovers potential communication problems. The majority of soloists are individual contributors (96%), and 4% of soloists are from middle management. These managers might face communication difficulties. We found an association between being an influencer and attitude toward a company's direction. There is a statistically significant 20% higher perception that the company is headed in the right direction among influencers compared to non-influencers (p < 0.05, Mann-Whitney test). Taken together, we demonstrate that considering social roles in the company might indicate both positive and negative aspects of organizational functioning that should be considered in data-driven decision-making.

Keywords: organizational network analysis, social roles, influencer, bridge, soloist

Procedia PDF Downloads 81

Authors: Guler Duru Asiret, Cemile Kutmec Yilmaz, Gulcan Bagcivan, Tugce Turten Kaymaz

Abstract:

This descriptive study was conducted to determine the relationship between caregiver burden and life satisfaction who give home care to elderly individuals. The sample was recruited from the internal medicine unit and palliative unit of a state hospital located in Turkey on June 2016-2017. The study sample consisted of 231 primary caregiver family member, who met the eligibility criteria and agreed to participate in the study. The inclusion criteria were as follows: inpatient’s caregiver, primary caregiver for at least 3 months, at least 18 years of age, no communication problem or mental disorder. Data were gathered using an Information Form prepared by the researchers based on previous literature, the Zarit Burden Interview (ZBI), and the Satisfaction with Life Scale (SWLS). The data were analyzed using IBM SPSS Statistics software version 20.0 (SPSS, Chicago, IL). The descriptive characteristics of the participant were analyzed using number, percentage, mean and standard deviation. The suitability of normal distribution of scale scores was analyzed using Kolmogorov-Smirnov and Shapiro-Wilk test. Relationships between scales were analyzed using Spearman’s rank-correlation coefficient. P values less than 0.05 were considered to be significant. The average age of the caregivers was 50.11±13.46 (mean±SD) years. Of the caregivers, 76.2% were women, 45% were primary school graduates, 89.2% were married, 38.1% were the daughters of their patients. Among these, 52.4% evaluated their income level to be good. Of them, 53.6% had been giving care less than 2 years. The patients’ average age was 77.1±8.0 years. Of the patients, 55.8% were women, 56.3% were illeterate, 70.6% were married, and 97.4% had at least one chronic disease. The mean Zarit Burden Interview score was 35.4±1.5 and the Satisfaction with Life Scale score was 20.6±6.8. A negative relationship was found between the patients’ score average on the ZBI, and on the SWLS (r= -0.438, p=0.000). The present study determined that the caregivers have a moderate caregiver burden and the life satisfaction. And the life satisfaction of caregivers decreased as their caregiver burden increase. In line with the results obtained from the research, it is recommended that to increase the effectiveness of discharge training, to arrange training and counseling programs for caregivers to cope with the problems they experienced, to monitor the caregivers at regular intervals and to provide necessary institutional support.

Keywords: caregiver burden, family caregivers, nurses, satisfaction

Procedia PDF Downloads 151

Authors: Michael Karnes, Alper Yilmaz

Abstract:

This paper presents the application of the omni-modeler towards pedestrian redetection. The pedestrian redetection task creates several challenges when applying deep neural networks (DNN) due to the variety of pedestrian appearance with camera position, the variety of environmental conditions, and the specificity required to recognize one pedestrian from another. DNNs require significant training sets and are not easily adapted for changes in class appearances or changes in the set of classes held in its knowledge domain. Pedestrian redetection requires an algorithm that can actively manage its knowledge domain as individuals move in and out of the scene, as well as learn individual appearances from a few frames of a video. The Omni-Modeler is a dynamically learning few-shot visual recognition algorithm developed for tasks with limited training data availability. The Omni-Modeler adapts the knowledge domain of pre-trained deep neural networks to novel concepts with a calculated localized language encoder. The Omni-Modeler knowledge domain is generated by creating a dynamic dictionary of concept definitions, which are directly updatable as new information becomes available. Query images are identified through nearest neighbor comparison to the learned object definitions. The study presented in this paper evaluates its performance in re-identifying individuals as they move through a scene in both single-camera and multi-camera tracking applications. The results demonstrate that the Omni-Modeler shows potential for across-camera view pedestrian redetection and is highly effective for single-camera redetection with a 93% accuracy across 30 individuals using 64 example images for each individual.

Keywords: dynamic learning, few-shot learning, pedestrian redetection, visual recognition

Procedia PDF Downloads 56

Authors: Mohammad Arabi

Abstract:

The development of advanced technologies in the field of signal processing and vibration analysis has enabled more accurate analysis and fault detection in electrical systems. This research investigates the application of temporal and frequency features in detecting faults in electric motor bearings, aiming to enhance fault detection accuracy and prevent unexpected failures. The use of methods such as deep learning algorithms and neural networks in this process can yield better results. The main objective of this research is to evaluate the efficiency and accuracy of methods based on temporal and frequency features in identifying faults in electric motor bearings to prevent sudden breakdowns and operational issues. Additionally, the feasibility of using techniques such as machine learning and optimization algorithms to improve the fault detection process is also considered. This research employed an experimental method and random sampling. Vibration signals were collected from electric motors under normal and faulty conditions. After standardizing the data, temporal and frequency features were extracted. These features were then analyzed using statistical methods such as analysis of variance (ANOVA) and t-tests, as well as machine learning algorithms like artificial neural networks and support vector machines (SVM). The results showed that using temporal and frequency features significantly improves the accuracy of fault detection in electric motor bearings. ANOVA indicated significant differences between normal and faulty signals. Additionally, t-tests confirmed statistically significant differences between the features extracted from normal and faulty signals. Machine learning algorithms such as neural networks and SVM also significantly increased detection accuracy, demonstrating high effectiveness in timely and accurate fault detection. This study demonstrates that using temporal and frequency features combined with machine learning algorithms can serve as an effective tool for detecting faults in electric motor bearings. This approach not only enhances fault detection accuracy but also simplifies and streamlines the detection process. However, challenges such as data standardization and the cost of implementing advanced monitoring systems must also be considered. Utilizing temporal and frequency features in fault detection of electric motor bearings, along with advanced machine learning methods, offers an effective solution for preventing failures and ensuring the operational health of electric motors. Given the promising results of this research, it is recommended that this technology be more widely adopted in industrial maintenance processes.

Keywords: electric motor, fault detection, frequency features, temporal features

Procedia PDF Downloads 19

Authors: Shubham Gogri, Lucia Florescu

Abstract:

Medical diagnosis and radiotherapy treatment planning using Computed Tomography (CT) rely on the quantitative accuracy and quality of the CT images. At the same time, requirements for CT imaging include reducing the radiation dose exposure to patients and minimizing scanning time. A solution to this is the sparse-view CT technique, based on a reduced number of projection views. This, however, introduces a new problem— the incomplete projection data results in lower quality of the reconstructed images. To tackle this issue, deep learning methods have been applied to enhance the quality of the sparse-view CT images. A first approach involved employing Mir-Net, a dedicated deep neural network designed for image enhancement. This showed promise, utilizing an intricate architecture comprising encoder and decoder networks, along with the incorporation of the Charbonnier Loss. However, this approach was computationally demanding. Subsequently, a specialized Generative Adversarial Network (GAN) architecture, rooted in the Pix2Pix framework, was implemented. This GAN framework involves a U-Net-based Generator and a Discriminator based on Convolutional Neural Networks. To bolster the GAN's performance, both Charbonnier and Wasserstein loss functions were introduced, collectively focusing on capturing minute details while ensuring training stability. The integration of the perceptual loss, calculated based on feature vectors extracted from the VGG16 network pretrained on the ImageNet dataset, further enhanced the network's ability to synthesize relevant images. A series of comprehensive experiments with clinical CT data were conducted, exploring various GAN loss functions, including Wasserstein, Charbonnier, and perceptual loss. The outcomes demonstrated significant image quality improvements, confirmed through pertinent metrics such as Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index (SSIM) between the corrected images and the ground truth. Furthermore, learning curves and qualitative comparisons added evidence of the enhanced image quality and the network's increased stability, while preserving pixel value intensity. The experiments underscored the potential of deep learning frameworks in enhancing the visual interpretation of CT scans, achieving outcomes with SSIM values close to one and PSNR values reaching up to 76.

Keywords: generative adversarial networks, sparse view computed tomography, CT image correction, Mir-Net

Procedia PDF Downloads 130

Authors: Emanuel Teixeira, Anderson Ramos, Marisa Lourenço, Fernando J. Velez, Jon M. Peha

Abstract:

This article discusses the benefit cost analysis aspects of millimetre wavebands (mmWaves) and Super High Frequency (SHF). The devaluation along the distance of the carrier-to-noise-plus-interference ratio with the coverage distance is assessed by considering two different path loss models, the two-slope urban micro Line-of-Sight (UMiLoS) for the SHF band and the modified Friis propagation model, for frequencies above 24 GHz. The equivalent supported throughput is estimated at the 5.62, 28, 38, 60 and 73 GHz frequency bands and the influence of carrier-to-noise-plus-interference ratio in the radio and network optimization process is explored. Mostly owing to the lessening caused by the behaviour of the two-slope propagation model for SHF band, the supported throughput at this band is higher than at the millimetre wavebands only for the longest cell lengths. The benefit cost analysis of these pico-cellular networks was analysed for regular cellular topologies, by considering the unlicensed spectrum. For shortest distances, we can distinguish an optimal of the revenue in percentage terms for values of the cell length, R ≈ 10 m for the millimeter wavebands and for longest distances an optimal of the revenue can be observed at R ≈ 550 m for the 5.62 GHz. It is possible to observe that, for the 5.62 GHz band, the profit is slightly inferior than for millimetre wavebands, for the shortest Rs, and starts to increase for cell lengths approximately equal to the ratio between the break-point distance and the co-channel reuse factor, achieving a maximum for values of R approximately equal to 550 m.

Keywords: millimetre wavebands, SHF band, SINR, cost benefit analysis, 5G

Procedia PDF Downloads 125

Authors: Monica Plechero

Abstract:

Many firms belonging to Italian industrial districts faced a crisis starting from 2000 and upsurging during 2008-2014. To remain competitive in the global market, these firms and their local systems need to renovate their traditional competitive advantages, strengthen their link with global flows of knowledge. This may be particularly relevant in sectors such as the mechatronics, that combine traditional knowledge domain with new knowledge domains (e.g. mechanics, electronics, and informatics). This sector is nowadays one of the key sectors within the so-called ‘smart specialization strategy’ that can lead part of the Italian traditional industry towards new economic developmental opportunities. This paper, by investigating the mechatronics district of the Veneto region, wants to shed new light on how firms of a local system can gain from the globalization of innovation and innovation networks. Methodologically, the paper relies on primary data collected through a survey targeting firms of the local system, as well as on a number of qualitative case studies. The relevant role of medium size companies in the district emerges as evident, as they have wider opportunities to be involved in different processes of globalization of innovation. Indeed, with respect to small companies, the size of medium firms allows them to exploit strategically international markets and globally distributed knowledge. Supporting medium firms’ global innovation strategies, and incentivizing their role as district gatekeepers, may strengthen the competitive capability of the local system and provide new opportunities to positively face global competition.

Keywords: global innovation network, industrial district, internationalization, innovation, mechatronics, Veneto region

Procedia PDF Downloads 212

Authors: Headman Hebe, Arnold Taringa

Abstract:

Background of the study: The benefits derived from Environmental and Sustainability Education (ESE) go beyond obtaining knowledge about the environment and the impact of human beings on the environment. Hence, it is sensible to expose learners to various resources that could enable meaningful environment-inclined pedagogy. The schoolgrounds, where they are utilised to promote ESE, benefit holistic learner development. However, empirical evidence, globally, suggests that young children’s contact with nature is declining due to urbanization, safety concerns by parents/guardians, and greater dependency on technology. Modern children spend much time on videogames and social media with very little time in the natural environment. Furthermore, national education departments in numerous countries have made tangible efforts to embed environmental and place-based learning to their school curricula. South Africa is one of those countries whose national school education curriculum advocates for ESE in pedagogy. Nevertheless, there is paucity of research conducted in South Africa on schoolgrounds as potential enablers of ESE and tools to foster a connection between youngsters and the natural environment. Accordingly, this study was essential as it seeks to determine the extent to which environmental learning is accommodated in pedagogy. Significantly, it investigates efforts made to use schoolgrounds for pedagogical purposes to connect children with the natural environment. Therefore, this study was conducted to investigate the accessibility and use of schoolgrounds for environment-inclined pedagogy in Natural and Social Sciences in two schools located in the Mpumalanga Province of South Africa. It tries to answer the question: To what extent are schoolgrounds used to promote environmental and sustainability education in the selected schools?The sub-questions: How do teachers and learners perceive the use of schoolgrounds for environmental and sustainability education activities? How does the organization of schoolgrounds offer opportunities for environmental education activities and accessibility for learners? Research method: This qualitative–interpretive case study used purposive and convenient sampling for participant selection. Forty-six respondents: 40 learners (twenty grade 7 learners per school), 2 school principals and 4 grade 7 participated in this study. Data collection tools were observations, interviews, audio-visual recordings and questionnaires while data analysis was done thematically. Major findings: The findings of the study point to: The lack of teacher training and infrastructure in the schoolgrounds and, no administrative support. Unclear curriculum guidelines on the use of schoolgrounds for ESE. The availability various elements in the schoolgrounds that could aid ESE activities. Learners denied access to certain parts of the schoolgrounds. Lack of time and curriculum demands constrain teachers from using schoolgrounds.

Keywords: affordances, environment and sustainability education, experiential learning, schoolgrounds

Procedia PDF Downloads 42

Authors: R. Geetha, J. Arunkumar, P. Gopal, D. Loganathan, K. Pavithra, C. Vikashini

Abstract:

Mobile Number Portability is an attempt to switch over from one network to another network facility for mobile based on applications. This facility is currently not available for mobile handsets. This application is intended to assist the mobile network and its service customers in understanding the criteria; this will serve as a universal set of requirements which must be met by the customers. This application helps the user's network portability. Accessing permission from the network provider to enable services to the user and utilizing the available network signals. It is enabling the user to make a temporary switch over to other network. The main aim of this research work is to adapt multiple networks at the time of no network coverage. It can be accessed at rural and geographical areas. This can be achieved by this mobile application. The application is capable of temporary switch over between various networks. With this application both the service provider and the network user are benefited. The service provider is benefited by charging a minimum cost for utilizing other network. It provides security in terms of password that is unique to avoid unauthorized users and to prevent loss of balance. The goal intended to be attained is a complete utilization of available network at significant situations and to provide feature that satisfy the customer needs. The temporary switch over is done to manage emergency calls when user is in rural or geographical area, where there will be a very low network coverage. Since people find it trend in using Android mobile, this application is designed as an Android applications, which can be freely downloaded and installed from Play store. In the current scenario, the service provider enables the user to change their network without shifting their mobile network. This application affords a clarification for users while they are jammed in a critical situation. This application is designed by using Android 4.2 and SQLite Version3.

Keywords: mobile number, random number, alarm, imei number, call

Procedia PDF Downloads 342

Authors: M. K. Wijayarathna, S. M. Buddika Harshanath

Abstract:

Monitoring student engagement has become a crucial part of the educational process and a reliable indicator of the capacity to retain information. As online learning classrooms are now more common these days, students' attention levels have become increasingly important, making it more difficult to check each student's concentration level in an online classroom setting. To profile student attention to various gradients of engagement, a study is a plan to conduct using machine learning models. Using a convolutional neural network, the findings and confidence score of the high accuracy model are obtained. In this research, convolutional neural networks are using to help discover essential emotions that are critical in defining various levels of participation. Students' attention levels were shown to be influenced by emotions such as calm, enjoyment, surprise, and fear. An improved virtual learning system was created as a result of these data, which allowed teachers to focus their support and advise on those students who needed it. Student participation has formed as a crucial component of the learning technique and a consistent predictor of a student's capacity to retain material in the classroom. Convolutional neural networks have a plan to implement the platform. As a preliminary step, a video of the pupil would be taken. In the end, researchers used a convolutional neural network utilizing the Keras toolkit to take pictures of the recordings. Two convolutional neural network methods are planned to use to determine the pupils' attention level. Finally, those predicted student attention level results plan to display on the graphical user interface of the System.

Keywords: HTML5, JavaScript, Python flask framework, AI, graphical user

Procedia PDF Downloads 78

Authors: Gunasekaran Raja, Ramkumar Jayaraman

Abstract:

In this paper, we consider CCL-N (Cooperative Cross Layer Network) topology based on the cross layer (both centralized and distributed) environment to form network communities. Various performance metrics related to the IEEE 802.16 networks are discussed to design CCL-N Topology. In CCL-N topology, nodes are classified as master nodes (Master Base Station [MBS]) and serving nodes (Relay Station [RS]). Nodes communities are organized based on the networking terminologies. Based on CCL-N Topology, various simulation analyses for both transparent and non-transparent relays are tabulated and throughput efficiency is calculated. Weighted load balancing problem plays a challenging role in IEEE 802.16 network. CoTS (Concurrent Transmission Scheduling) Scheme is formulated in terms of three aspects – transmission mechanism based on identical communities, different communities and identical node communities. CoTS scheme helps in identifying the weighted load balancing problem. Based on the analytical results, modularity value is inversely proportional to that of the error value. The modularity value plays a key role in solving the CoTS problem based on hop count. The transmission mechanism for identical node community has no impact since modularity value is same for all the network groups. In this paper three aspects of communities based on the modularity value which helps in solving the problem of weighted load balancing and CoTS are discussed.

Keywords: cross layer network topology, concurrent scheduling, modularity value, network communities and weighted load balancing

Procedia PDF Downloads 241

Authors: Demet Ozgur-Unluakin, Busenur Turkali, Ayse Karacaorenli

Abstract:

Maintenance strategies can be classified into two types, which are reactive and proactive, with respect to the time of the failure and maintenance. If the maintenance activity is done after a breakdown, it is called reactive maintenance. On the other hand, proactive maintenance, which is further divided as preventive and predictive, focuses on maintaining components before a failure occurs to prevent expensive halts. Recently, the number of interacting components in a system has increased rapidly and therefore, the structure of the systems have become more complex. This situation has made it difficult to provide the right maintenance decisions. Herewith, determining effective decisions has played a significant role. In multi-component systems, many methodologies and strategies can be applied when a component or a system has already broken down or when it is desired to identify and avoid proactively defects that could lead to future failure. This study focuses on the comparison of various maintenance strategies on a multi-component dynamic system. Components in the system are hidden, although there exists partial observability to the decision maker and they deteriorate in time. Several predefined policies under corrective, preventive and predictive maintenance strategies are considered to minimize the total maintenance cost in a planning horizon. The policies are simulated via Dynamic Bayesian Networks on a multi-component system with different policy parameters and cost scenarios, and their performances are evaluated. Results show that when the difference between the corrective and proactive maintenance cost is low, none of the proactive maintenance policies is significantly better than the corrective maintenance. However, when the difference is increased, at least one policy parameter for each proactive maintenance strategy gives significantly lower cost than the corrective maintenance.

Keywords: decision making, dynamic Bayesian networks, maintenance, multi-component systems, reliability

Procedia PDF Downloads 111

Authors: Johannes Dregger, Michael Schmidt, Christian Prasse, Michael ten Hompel

Abstract:

Most of the today’s companies face increasing need to operate efficiently. Driven by global trends like shorter product cycles, mass customization and the rising speed of delivery, manufacturing value chains are becoming more and more distributed. Manufacturing processes are becoming highly integrated, e.g. 3D printing. All these changes are affecting companies´ organization. They are leading towards individual, small scale, and ad-hoc logistics processes and structures, and finally, towards a significant increase in the importance of logistics itself since traditional value chains transform into agile value networks. In the past logistics has been following manufacturing but in the future industry, this role allocation might change. With this increase in the logistics practice of companies and businesses, the relevance of logistics research as the methodological foundation of logistics networks and processes is gaining importance. Logistics research is evolving into a central and highly interdisciplinary science for the future industry. Using the example of Germany, this paper discusses ways to establish logistics as a central scientific discipline of the future industry. About three million people work in the logistics sector in Germany. Only automotive and retail industry have more employees. Even though there is a bunch of logistics degree programs at more than 100 institutions of higher education, a common understanding of logistics as a research discipline is missing. In this paper an innovative approach will be presented, including; identified perspectives on logistics, such as process orientation, IT orientation or employees orientation, relevant scientific disciplines for logistics science, a concept for interdisciplinary research approaches to unify the perspectives of the different scientific disciplines on logistics and the methodological base of logistics science.

Keywords: logistics, logistics science, logistics management, future challenges

Procedia PDF Downloads 293

Authors: Alaa Allakany, Koji Okamura

Abstract:

Multicast technology is an efficient and scalable technology for data distribution in order to optimize network resources. However, in the IP network, the responsibility for management of multicast groups is distributed among network routers, which causes some limitations such as delays in processing group events, high bandwidth consumption and redundant tree calculation. Software Defined Networking (SDN) represented by OpenFlow presented as a solution for many problems, in SDN the control plane and data plane are separated by shifting the control and management to a remote centralized controller, and the routers are used as a forwarder only. In this paper we will proposed fast switching mechanism for solving the problem of link failure in multicast tree based on Tabu Search heuristic algorithm and modifying the functions of OpenFlow switch to fasts switch to the pack up sub tree rather than sending to the controller. In this work we will implement multicasting OpenFlow controller, this centralized controller is a core part in our multicasting approach, which is responsible for 1- constructing the multicast tree, 2- handling the multicast group events and multicast state maintenance. And finally modifying OpenFlow switch functions for fasts switch to pack up paths. Forwarders, forward the multicast packet based on multicast routing entries which were generated by the centralized controller. Tabu search will be used as heuristic algorithm for construction near optimum multicast tree and maintain multicast tree to still near optimum in case of join or leave any members from multicast group (group events).

Keywords: multicast tree, software define networks, tabu search, OpenFlow

Procedia PDF Downloads 243

Authors: Chapman Eze Nnadozie

Abstract:

Cloud computing, a technology that is made possible through virtualization within networks represents a shift from the traditional ownership of infrastructure and other resources by distinct organization to a more scalable pattern in which computer resources are rented online to organizations on either as a pay-as-you-use basis or by subscription. In other words, cloud computing entails the renting of computing resources (such as storage space, memory, servers, applications, networks, etc.) by a third party to its clients on a pay-as-go basis. It is a new innovative technology that is globally embraced because of its renowned benefits, profound of which is its cost effectiveness on the part of organizations engaged with its services. In Nigeria, the services are provided either directly to companies mostly by the key IT players such as Microsoft, IBM, and Google; or in partnership with some other players such as Infoware, Descasio, and Sunnet. This action enables organizations to rent IT resources on a pay-as-you-go basis thereby salvaging them from wastages accruable on acquisition and maintenance of IT resources such as ownership of a separate data centre. This paper intends to appraise the challenges of cloud computing adoption in Nigeria, bearing in mind the country’s peculiarities’ in terms of infrastructural development. The methodologies used in this paper include the use of research questionnaires, formulated hypothesis, and the testing of the formulated hypothesis. The major findings of this paper include the fact that there are some addressable challenges to the adoption of cloud computing in Nigeria. Furthermore, the country will gain significantly if the challenges especially in the area of infrastructural development are well addressed. This is because the research established the fact that there are significant gains derivable by the adoption of cloud computing by organizations in Nigeria. However, these challenges can be overturned by concerted efforts in the part of government and other stakeholders.

Keywords: cloud computing, data centre, infrastructure, it resources, virtualization

Procedia PDF Downloads 331

Authors: Muhammet Baldan, Emel Timuçin

Abstract:

Molecular solubility prediction plays a crucial role in various fields, such as drug discovery, environmental science, and material science. In this study, we compare the performance of five machine learning algorithms—linear regression, support vector machines (SVM), random forests, gradient boosting machines (GBM), and neural networks—for predicting molecular solubility using the AqSolDB dataset. The dataset consists of 9981 data points with their corresponding solubility values. MACCS keys (166 bits), RDKit properties (20 properties), and structural properties(3) features are extracted for every smile representation in the dataset. A total of 189 features were used for training and testing for every molecule. Each algorithm is trained on a subset of the dataset and evaluated using metrics accuracy scores. Additionally, computational time for training and testing is recorded to assess the efficiency of each algorithm. Our results demonstrate that random forest model outperformed other algorithms in terms of predictive accuracy, achieving an 0.93 accuracy score. Gradient boosting machines and neural networks also exhibit strong performance, closely followed by support vector machines. Linear regression, while simpler in nature, demonstrates competitive performance but with slightly higher errors compared to ensemble methods. Overall, this study provides valuable insights into the performance of machine learning algorithms for molecular solubility prediction, highlighting the importance of algorithm selection in achieving accurate and efficient predictions in practical applications.

Keywords: random forest, machine learning, comparison, feature extraction

Procedia PDF Downloads 16

Authors: Yasin Tadayonrad

Abstract:

Most of the supply chain networks have many nodes starting from the suppliers’ side up to the customers’ side that each node sends/receives the raw materials/products from/to the other nodes. One of the major concerns in this kind of supply chain network is finding the best schedule for loading /unloading the shipments through the whole network by which all the constraints in the source and destination nodes are met and all the shipments are delivered on time. One of the main constraints in this problem is loading/unloading capacity in each source/ destination node at each time slot (e.g., per week/day/hour). Because of the different characteristics of different products/groups of products, the capacity of each node might differ based on each group of products. In most supply chain networks (especially in the Fast-moving consumer goods industry), there are different planners/planning teams working separately in different nodes to determine the loading/unloading timeslots in source/destination nodes to send/receive the shipments. In this paper, a mathematical problem has been proposed to find the best timeslots for loading/unloading the shipments minimizing the overall delays subject to respecting the capacity of loading/unloading of each node, the required delivery date of each shipment (considering the lead-times), and working-days of each node. This model was implemented on python and solved using Python-MIP on a sample data set. Finally, the idea of a heuristic algorithm has been proposed as a way of improving the solution method that helps to implement the model on larger data sets in real business cases, including more nodes and shipments.

Keywords: supply chain management, transportation, multiple-multiple network, timeslots management, mathematical modeling, mixed integer programming

Procedia PDF Downloads 79

Authors: A. Alzahrani, G. Arnold, W. Wang

Abstract:

Background: Stroke commonly occurs in middle-aged and elderly populations, and the diagnosis of early stroke is still difficult. Patients who have suffered a stroke have different balance and gait patterns from healthy people. Advanced techniques of motion analysis have been routinely used in the clinical assessment of cerebral palsy. However, so far, little research has been done on the direct diagnosis of early stroke patients using motion analysis. Objectives: The aim of this study was to investigate whether patients with stroke have different balance and gait from healthy people and which biomechanical parameters could be used to predict and diagnose potential patients who are at a potential risk to stroke. Methods: Thirteen patients with stroke were recruited as subjects whose gait and balance was analysed. Twenty normal subjects at the matched age participated in this study as a control group. All subjects’ gait and balance were collected using Vicon Nexus® to obtain the gait parameters, kinetic, and kinematic parameters of the hip, knee, and ankle joints in three planes of both limbs. Participants stood on force platforms to perform a single leg balance test. Then, they were asked to walk along a 10 m walkway at their comfortable speed. Participants performed 6 trials of single-leg balance for each side and 10 trials of walking. From the recorded trials, three good ones were analysed using the Vicon Plug-in-Gait model to obtain gait parameters, e.g., walking speed, cadence, stride length, and joint parameters, e.g., joint angle, force, moments, etc. Result: The temporal-spatial variables of Stroke subjects were compared with the healthy subjects; it was found that there was a significant difference (p < 0.05) between the groups. The step length, speed, cadence were lower in stroke subjects as compared to the healthy groups. The stroke patients group showed significantly decreased in gait speed (mean and SD: 0.85 ± 0.33 m/s), cadence ( 96.71 ± 16.14 step/min), and step length (0.509 ± 017 m) in compared to healthy people group whereas the gait speed was 1.2 ± 0.11 m/s, cadence 112 ± 8.33 step/min, and step length 0.648 ± 0.43 m. Moreover, it was observed that patients with stroke have significant differences in the ankle, hip, and knee joints’ kinematics in the sagittal and coronal planes. Also, the result showed that there was a significant difference between groups in the single-leg balance test, e.g., maintaining single-leg stance time in the stroke patients showed shorter duration (5.97 ± 6.36 s) in compared to healthy people group (14.36 ± 10.20 s). Conclusion: Our result showed that there are significantly differences between stroke patients and healthy subjects in the various aspects of gait analysis and balance test, as a consequences of these findings some of the biomechanical parameters such as joints kinematics, gait parameters, and single-leg stance balance test could be used in clinical practice to predict and diagnose potential patients who are at a high risk of further stroke.

Keywords: gait analysis, kinetics, kinematics, single-leg stance, Stroke

Procedia PDF Downloads 126

Authors: D. Y. Fong, C. Y. Kuo, Y. T. Chiang, W. C. Lin

Abstract:

Objectives: Qigong is an ancient Chinese practice in pursuit of a healthier body and a more peaceful mindset. It emphasizes on the restoration of vital energy (Qi) in body, mind, and spirit. The practice is the combination of gentle movements and mild breathing which help the doers reach the condition of tranquility. On account of the features of Qigong, first, we use cross-sectional methodology to compare the differences among the varied levels of Qigong practitioners on cognitive function with event-related potential (ERP) and electroencephalography (EEG). Second, we use the longitudinal methodology to explore the effects on the Qigong trainees for pretest and posttest on ERP and EEG. Current study adopts Attentional Network Test (ANT) task to examine the participants’ cognitive function, and aging-related researches demonstrated a declined tread on the cognition in older adults and exercise might ameliorate the deterioration. Qigong exercise integrates physical posture (muscle strength), breathing technique (aerobic ability) and focused intention (attention) that researchers hypothesize it might improve the cognitive function in aging adults. Method: Sixty participants were involved in this study, including 20 young adults (21.65±2.41 y) with normal physical activity (YA), 20 Qigong experts (60.69 ± 12.42 y) with over 7 years Qigong practice experience (QE), and 20 normal and healthy adults (52.90±12.37 y) with no Qigong practice experience as experimental group (EG). The EG participants took Qigong classes 2 times a week and 2 hours per time for 24 weeks with the purpose of examining the effect of Qigong intervention on cognitive function. ANT tasks (alert network, orient network, and executive control) were adopted to evaluate participants’ cognitive function via ERP’s P300 components and P300 amplitude topography. Results: Behavioral data: 1.The reaction time (RT) of YA is faster than the other two groups, and EG was faster than QE in the cue and flanker conditions of ANT task. 2. The RT of posttest was faster than pretest in EG in the cue and flanker conditions. 3. No difference among the three groups on orient, alert, and execute control networks. ERP data: 1. P300 amplitude detection in QE was larger than EG at Fz electrode in orient, alert, and execute control networks. 2. P300 amplitude in EG was larger at pretest than posttest on the orient network. 3. P300 Latency revealed no difference among the three groups in the three networks. Conclusion: Taken together these findings, they provide neuro-electrical evidence that older adults involved in Qigong practice may develop a more overall compensatory mechanism and also benefit the performance of behavior.

Keywords: Qigong, cognitive function, aging, event-related potential (ERP)

Procedia PDF Downloads 381

Authors: Xiaolei Hu, Enrico Ferrera, Riccardo Tomasi, Claudio Pastrone

Abstract:

Load Forecasting plays a key role in making today's and future's Smart Energy Grids sustainable and reliable. Accurate power consumption prediction allows utilities to organize in advance their resources or to execute Demand Response strategies more effectively, which enables several features such as higher sustainability, better quality of service, and affordable electricity tariffs. It is easy yet effective to apply Load Forecasting at larger geographic scale, i.e. Smart Micro Grids, wherein the lower available grid flexibility makes accurate prediction more critical in Demand Response applications. This paper analyses the application of short-term load forecasting in a concrete scenario, proposed within the EU-funded GreenCom project, which collect load data from single loads and households belonging to a Smart Micro Grid. Three short-term load forecasting techniques, i.e. linear regression, artificial neural networks, and radial basis function network, are considered, compared, and evaluated through absolute forecast errors and training time. The influence of weather conditions in Load Forecasting is also evaluated. A new definition of Gain is introduced in this paper, which innovatively serves as an indicator of short-term prediction capabilities of time spam consistency. Two models, 24- and 1-hour-ahead forecasting, are built to comprehensively compare these three techniques.

Keywords: short-term load forecasting, smart micro grid, linear regression, artificial neural networks, radial basis function network, gain

Procedia PDF Downloads 444

Authors: Nur Bella Octoria Bella, Ayomi Dita Rarasati

Abstract:

DKI Jakarta is Indonesia's capital city with the 10th highest congestion rate in the world based on the 2019 traffic index. Other than that based on World Air Quality Report in 2019 showed DKI Jakarta's air pollutant concentrate 49.4 µg and the 5th highest air pollutant in the world. In the urban city nowadays, the mobility rate is high enough and the efficiency for sustainability assessment in transport infrastructure development is needed. This efficiency is the important key for sustainable infrastructure development. DKI Jakarta is nowadays in the process of constructing the railway infrastructure to support the transportation system. The problems appearing are the railway infrastructure networks and the service in DKI Jakarta already planned based on sustainability factors or not. Therefore, the aim of this research is to make the evaluation of railways infrastructure networks performance and services in DKI Jakarta regards on the railway sustainability key factors. Further, this evaluation will be used to make the railway sustainability assessment framework and to offer some of the alternative solutions to improve railway transportation sustainability in DKI Jakarta. Firstly a very detailed literature review of papers that have focused on railway sustainability factors and their improvements of railway sustainability, published in the scientific journal in the period 2011 until 2021. Regarding the sustainability factors from the literature review, further, it is used to assess the current condition of railway infrastructure in DKI Jakarta. The evaluation will be using a Likert rate questionnaire and directed to the transportation railway expert and the passenger. Furthermore, the mapping and evaluation rate based on the sustainability factors will be compared to the effect factors using the Analytical Hierarchical Process (AHP). This research offers the network's performance and service rate impact on the sustainability aspect and the passenger willingness for using the rail public transportation in DKI Jakarta.

Keywords: transportation sustainability, railway transportation, sustainability, DKI Jakarta

Procedia PDF Downloads 144

Authors: Shraddha Jain Sharma, Ratnalata Gupta

Abstract:

Predicting stock movements in the financial market is both difficult and rewarding. Analysts and academics are increasingly using advanced approaches such as machine learning techniques to anticipate stock price patterns, thanks to the expanding capacity of computing and the recent advent of graphics processing units and tensor processing units. Stock market prediction is a type of time series prediction that is incredibly difficult to do since stock prices are influenced by a variety of financial, socioeconomic, and political factors. Furthermore, even minor mistakes in stock market price forecasts can result in significant losses for companies that employ the findings of stock market price prediction for financial analysis and investment. Soft computing techniques are increasingly being employed for stock market prediction due to their better accuracy than traditional statistical methodologies. The proposed research looks at the need for soft computing techniques in stock market prediction, the numerous soft computing approaches that are important to the field, past work in the area with their prominent features, and the significant problems or issue domain that the area involves. For constructing a predictive model, the major focus is on neural networks and fuzzy logic. The stock market is extremely unpredictable, and it is unquestionably tough to correctly predict based on certain characteristics. This study provides a complete overview of the numerous strategies investigated for high accuracy prediction, with a focus on the most important characteristics.

Keywords: stock market prediction, artificial intelligence, artificial neural networks, fuzzy logic, accuracy, deep learning, machine learning, stock price, trading volume

Procedia PDF Downloads 66

Authors: Teja Sai Puligadda

Abstract:

Satellite images in detail can serve an important role in the geographic study. Quantitative and qualitative information provided by the satellite and remote sensing images minimizes the complexity of work and time. Data/images are captured at regular intervals by satellite remote sensing systems, and the amount of data collected is often enormous, and it expands rapidly as technology develops. Interpreting remote sensing images, geographic data mining, and researching distinct vegetation types such as agricultural and forests are all part of satellite image categorization. One of the biggest challenge data scientists faces while classifying satellite images is finding the best suitable classification algorithms based on the available that could able to classify images with utmost accuracy. In order to categorize satellite images, which is difficult due to the sheer volume of data, many academics are turning to deep learning machine algorithms. As, the CNN algorithm gives high accuracy in image recognition problems and automatically detects the important features without any human supervision and the ANN algorithm stores information on the entire network (Abhishek Gupta., 2020), these two deep learning algorithms have been used for satellite image classification. This project focuses on remote sensing through Deep Neural Networks i.e., ANN and CNN with Deep Sat (SAT-4) Airborne dataset for classifying images. Thus, in this project of classifying satellite images, the algorithms ANN and CNN are implemented, evaluated & compared and the performance is analyzed through evaluation metrics such as Accuracy and Loss. Additionally, the Neural Network algorithm which gives the lowest bias and lowest variance in solving multi-class satellite image classification is analyzed.

Keywords: artificial neural network, convolutional neural network, remote sensing, accuracy, loss

Procedia PDF Downloads 137

Authors: Alex Fedoseyev

Abstract:

This study investigates a generalized hydrodynamic equation (GHE) simplified model for the simulation of turbulent flow over a two-dimensional backward-facing step (BFS) at Reynolds number Re=132000. The GHE were derived from the generalized Boltzmann equation (GBE). GBE was obtained by first principles from the chain of Bogolubov kinetic equations and considers particles of finite dimensions. The GHE has additional terms, temporal and spatial fluctuations, compared to the Navier-Stokes equations (NSE). These terms have a timescale multiplier τ, and the GHE becomes the NSE when $\tau$ is zero. The nondimensional τ is a product of the Reynolds number and the squared length scale ratio, τ=Re*(l/L)², where l is the apparent Kolmogorov length scale, and L is a hydrodynamic length scale. The BFS flow modeling results obtained by 2D calculations cannot match the experimental data for Re>450. One or two additional equations are required for the turbulence model to be added to the NSE, which typically has two to five parameters to be tuned for specific problems. It is shown that the GHE does not require an additional turbulence model, whereas the turbulent velocity results are in good agreement with the experimental results. A review of several studies on the simulation of flow over the BFS from 1980 to 2023 is provided. Most of these studies used different turbulence models when Re>1000. In this study, the 2D turbulent flow over a BFS with height H=L/3 (where L is the channel height) at Reynolds number Re=132000 was investigated using numerical solutions of the GHE (by a finite-element method) and compared to the solutions from the Navier-Stokes equations, k–ε turbulence model, and experimental results. The comparison included the velocity profiles at X/L=5.33 (near the end of the recirculation zone, available from the experiment), recirculation zone length, and velocity flow field. The mean velocity of NSE was obtained by averaging the solution over the number of time steps. The solution with a standard k −ε model shows a velocity profile at X/L=5.33, which has no backward flow. A standard k−ε model underpredicts the experimental recirculation zone length X/L=7.0∓0.5 by a substantial amount of 20-25%, and a more sophisticated turbulence model is needed for this problem. The obtained data confirm that the GHE results are in good agreement with the experimental results for turbulent flow over two-dimensional BFS. A turbulence model was not required in this case. The computations were stable. The solution time for the GHE is the same or less than that for the NSE and significantly less than that for the NSE with the turbulence model. The proposed approach was limited to 2D and only one Reynolds number. Further work will extend this approach to 3D flow and a higher Re.

Keywords: backward-facing step, comparison with experimental data, generalized hydrodynamic equations, separation, reattachment, turbulent flow

Procedia PDF Downloads 40

Authors: Kamal KC, Zhendong Yin, Dasen Li, Zhilu Wu

Abstract:

Intra-plant disease classification based on leaf images is a challenging computer vision task due to similarities in texture, color, and shape of leaves with a slight variation of leaf spot; and external environmental changes such as lighting and background noises. Deep convolutional neural network (DCNN) has proven to be an effective tool for binary classification. In this paper, two methods for binary classification of diseased plant leaves using DCNN are presented; model created from scratch and transfer learning. Our main contribution is a thorough evaluation of 4 networks created from scratch and transfer learning of 5 pre-trained models. Training and testing of these models were performed on a plant leaf images dataset belonging to 16 distinct classes, containing a total of 22,265 images from 8 different plants, consisting of a pair of healthy and diseased leaves. We introduce a deep CNN model, Optimized MobileNet. This model with depthwise separable CNN as a building block attained an average test accuracy of 99.77%. We also present a fine-tuning method by introducing the concept of a convolutional block, which is a collection of different deep neural layers. Fine-tuned models proved to be efficient in terms of accuracy and computational cost. Fine-tuned MobileNet achieved an average test accuracy of 99.89% on 8 pairs of [healthy, diseased] leaf ImageSet.

Keywords: deep convolution neural network, depthwise separable convolution, fine-grained classification, MobileNet, plant disease, transfer learning

Procedia PDF Downloads 166

Authors: Adenike Adesanya, Nonthaphat Wong, Xiang-Yun Lan, Shea Ping Yip, Chien-Ling Huang

Abstract:

Background: The most challenging mutation of the oncokinase BCR-ABL protein T315I, which is commonly known as the “gatekeeper” mutation and is notorious for its strong resistance to almost all tyrosine kinase inhibitors (TKIs), especially imatinib. Therefore, this study aims to identify T315I-dependent downstream microRNA (miRNA) pathways associated with drug resistance in chronic myeloid leukemia (CML) for prognostic and therapeutic purposes. Methods: T315I-carrying K562 cell clones (K562-T315I) were generated by the CRISPR-Cas9 system. Imatinib-treated K562-T315I cells were subjected to small RNA library preparation and next-generation sequencing. Putative lncRNA-miRNA-mRNA networks were analyzed with (i) DESeq2 to extract differentially expressed miRNAs, using Padj value of 0.05 as cut-off, (ii) STarMir to obtain potential miRNA response element (MRE) binding sites of selected miRNAs on lncRNA H19, (iii) miRDB, miRTarbase, and TargetScan to predict mRNA targets of selected miRNAs, (iv) IntaRNA to obtain putative interactions between H19 and the predicted mRNAs, (v) Cytoscape to visualize putative networks, and (vi) several pathway analysis platforms – Enrichr, PANTHER and ShinyGO for pathway enrichment analysis. Moreover, mitochondria isolation and transcript quantification were adopted to determine the new mechanism involved in T315I-mediated resistance of CML treatment. Results: Verification of the CRISPR-mediated mutagenesis with digital droplet PCR detected the mutation abundance of ≥80%. Further validation showed the viability of ≥90% by cell viability assay, and intense phosphorylated CRKL protein band being detected with no observable change for BCR-ABL and c-ABL protein expressions by Western blot. As reported by several investigations into hematological malignancies, we determined a 7-fold increase of H19 expression in K562-T315I cells. After imatinib treatment, a 9-fold increment was observed. DESeq2 revealed 171 miRNAs were differentially expressed K562-T315I, 112 out of these miRNAs were identified to have MRE binding regions on H19, and 26 out of the 112 miRNAs were significantly downregulated. Adopting the seed-sequence analysis of these identified miRNAs, we obtained 167 mRNAs. 6 hub miRNAs (hsa-let-7b-5p, hsa-let-7e-5p, hsa-miR-125a-5p, hsa-miR-129-5p, and hsa-miR-372-3p) and 25 predicted genes were identified after constructing hub miRNA-target gene network. These targets demonstrated putative interactions with H19 lncRNA and were mostly enriched in pathways related to cell proliferation, senescence, gene silencing, and pluripotency of stem cells. Further experimental findings have also shown the up-regulation of mitochondrial transcript and lncRNA MALAT1 contributing to the lncRNA-miRNA-mRNA networks induced by BCR-ABL T315I mutation. Conclusions: Our results have indicated that lncRNA-miRNA regulators play a crucial role not only in leukemogenesis but also in drug resistance, considering the significant dysregulation and interactions in the K562-T315I cell model generated by CRISPR-Cas9. In silico analysis has further shown that lncRNAs H19 and MALAT1 bear several complementary miRNA sites. This implies that they could serve as a sponge, hence sequestering the activity of the target miRNAs.

Keywords: chronic myeloid leukemia, imatinib resistance, lncRNA-miRNA-mRNA, T315I mutation

Procedia PDF Downloads 137

Authors: Usama Abdulwahab, Mohammed Wahab

Abstract:

This paper presents the application of uncapacitated facility location problems (UFLP) and 1-median problems to support decision making in blood supply chain networks. A plethora of factors make blood supply-chain networks a complex, yet vital problem for the regional blood bank. These factors are rapidly increasing demand; criticality of the product; strict storage and handling requirements; and the vastness of the theater of operations. As in the UFLP, facilities can be opened at any of $m$ predefined locations with given fixed costs. Clients have to be allocated to the open facilities. In classical location models, the allocation cost is the distance between a client and an open facility. In this model, the costs are the allocation cost, transportation costs, and inventory costs. In order to address this problem the median algorithm is used to analyze inventory, evaluate supply chain status, monitor performance metrics at different levels of granularity, and detect potential problems and opportunities for improvement. The Euclidean distance data for some Ontario cities (demand nodes) are used to test the developed algorithm. Sitation software, lagrangian relaxation algorithm, and branch and bound heuristics are used to solve this model. Computational experiments confirm the efficiency of the proposed approach. Compared to the existing modeling and solution methods, the median algorithm approach not only provides a more general modeling framework but also leads to efficient solution times in general.

Keywords: approximate dynamic programming, facility location, perishable product, inventory model, blood platelet, P-median problem

Procedia PDF Downloads 488

Authors: Ethan James

Abstract:

Over one billion people worldwide suffer from some level of vision loss or blindness as a result of progressive retinal diseases. Many patients, particularly in developing areas, are incorrectly diagnosed or undiagnosed whatsoever due to unconventional diagnostic tools and screening methods. Artificial intelligence (AI) based on deep learning (DL) convolutional neural networks (CNN) have recently gained a high interest in ophthalmology for its computer-imaging diagnosis, disease prognosis, and risk assessment. Optical coherence tomography (OCT) is a popular imaging technique used to capture high-resolution cross-sections of retinas. In ophthalmology, DL has been applied to fundus photographs, optical coherence tomography, and visual fields, achieving robust classification performance in the detection of various retinal diseases including macular degeneration, diabetic retinopathy, and retinitis pigmentosa. However, there is no complete diagnostic model to analyze these retinal images that provide a diagnostic accuracy above 90%. Thus, the purpose of this project was to develop an AI model that utilizes machine learning techniques to automatically diagnose specific retinal diseases from OCT scans. The algorithm consists of neural network architecture that was trained from a dataset of over 20,000 real-world OCT images to train the robust model to utilize residual neural networks with cyclic pooling. This DL model can ultimately aid ophthalmologists in diagnosing patients with these retinal diseases more quickly and more accurately, therefore facilitating earlier treatment, which results in improved post-treatment outcomes.

Keywords: artificial intelligence, deep learning, imaging, medical devices, ophthalmic devices, ophthalmology, retina

Procedia PDF Downloads 157

Authors: Flavio Cannavo

Abstract:

Continuous evaluation of the status of potentially hazardous volcanos plays a key role for civil protection purposes. The importance of monitoring volcanic activity, especially for energetic paroxysms that usually come with tephra emissions, is crucial not only for exposures to the local population but also for airline traffic. Presently, real-time surveillance of most volcanoes worldwide is essentially delegated to one or more human experts in volcanology, who interpret data coming from different kind of monitoring networks. Unfavorably, the high nonlinearity of the complex and coupled volcanic dynamics leads to a large variety of different volcanic behaviors. Moreover, continuously measured parameters (e.g. seismic, deformation, infrasonic and geochemical signals) are often not able to fully explain the ongoing phenomenon, thus making the fast volcano state assessment a very puzzling task for the personnel on duty at the control rooms. With the aim of aiding the personnel on duty in volcano surveillance, here we introduce a system based on an ensemble of data-driven classifiers to infer automatically the ongoing volcano status from all the available different kind of measurements. The system consists of a heterogeneous set of independent classifiers, each one built with its own data and algorithm. Each classifier gives an output about the volcanic status. The ensemble technique allows weighting the single classifier output to combine all the classifications into a single status that maximizes the performance. We tested the model on the Mt. Etna (Italy) case study by considering a long record of multivariate data from 2011 to 2015 and cross-validated it. Results indicate that the proposed model is effective and of great power for decision-making purposes.

Keywords: Bayesian networks, expert system, mount Etna, volcano monitoring

Procedia PDF Downloads 226

Authors: Zhongmin Wang, Wudong Fan, Hengshan Zhang, Yimin Zhou

Abstract:

In data-driven prognostic methods, the prediction accuracy of the estimation for remaining useful life of bearings mainly depends on the performance of health indicators, which are usually fused some statistical features extracted from vibrating signals. However, the existing health indicators have the following two drawbacks: (1) The differnet ranges of the statistical features have the different contributions to construct the health indicators, the expert knowledge is required to extract the features. (2) When convolutional neural networks are utilized to tackle time-frequency features of signals, the time-series of signals are not considered. To overcome these drawbacks, in this study, the method combining convolutional neural network with gated recurrent unit is proposed to extract the time-frequency image features. The extracted features are utilized to construct health indicator and predict remaining useful life of bearings. First, original signals are converted into time-frequency images by using continuous wavelet transform so as to form the original feature sets. Second, with convolutional and pooling layers of convolutional neural networks, the most sensitive features of time-frequency images are selected from the original feature sets. Finally, these selected features are fed into the gated recurrent unit to construct the health indicator. The results state that the proposed method shows the enhance performance than the related studies which have used the same bearing dataset provided by PRONOSTIA.

Keywords: continuous wavelet transform, convolution neural net-work, gated recurrent unit, health indicators, remaining useful life

Procedia PDF Downloads 111