Search results for: forecasting accuracy
732 Document-level Sentiment Analysis: An Exploratory Case Study of Low-resource Language Urdu
Authors: Ammarah Irum, Muhammad Ali Tahir
Abstract:
Document-level sentiment analysis in Urdu is a challenging Natural Language Processing (NLP) task due to the difficulty of working with lengthy texts in a language with constrained resources. Deep learning models, which are complex neural network architectures, are well-suited to text-based applications in addition to data formats like audio, image, and video. To investigate the potential of deep learning for Urdu sentiment analysis, we implemented five different deep learning models, including Bidirectional Long Short Term Memory (BiLSTM), Convolutional Neural Network (CNN), Convolutional Neural Network with Bidirectional Long Short Term Memory (CNN-BiLSTM), and Bidirectional Encoder Representation from Transformer (BERT). In this study, we developed a hybrid deep learning model called BiLSTM-Single Layer Multi Filter Convolutional Neural Network (BiLSTM-SLMFCNN) by fusing BiLSTM and CNN architecture. The proposed and baseline techniques are applied on Urdu Customer Support data set and IMDB Urdu movie review data set by using pre-trained Urdu word embedding that are suitable for sentiment analysis at the document level. Results of these techniques are evaluated and our proposed model outperforms all other deep learning techniques for Urdu sentiment analysis. BiLSTM-SLMFCNN outperformed the baseline deep learning models and achieved 83%, 79%, 83% and 94% accuracy on small, medium and large sized IMDB Urdu movie review data set and Urdu Customer Support data set respectively.Keywords: urdu sentiment analysis, deep learning, natural language processing, opinion mining, low-resource language
Procedia PDF Downloads 63731 Predicting Match Outcomes in Team Sport via Machine Learning: Evidence from National Basketball Association
Authors: Jacky Liu
Abstract:
This paper develops a team sports outcome prediction system with potential for wide-ranging applications across various disciplines. Despite significant advancements in predictive analytics, existing studies in sports outcome predictions possess considerable limitations, including insufficient feature engineering and underutilization of advanced machine learning techniques, among others. To address these issues, we extend the Sports Cross Industry Standard Process for Data Mining (SRP-CRISP-DM) framework and propose a unique, comprehensive predictive system, using National Basketball Association (NBA) data as an example to test this extended framework. Our approach follows a holistic methodology in feature engineering, employing both Time Series and Non-Time Series Data, as well as conducting Explanatory Data Analysis and Feature Selection. Furthermore, we contribute to the discourse on target variable choice in team sports outcome prediction, asserting that point spread prediction yields higher profits as opposed to game-winner predictions. Using machine learning algorithms, particularly XGBoost, results in a significant improvement in predictive accuracy of team sports outcomes. Applied to point spread betting strategies, it offers an astounding annual return of approximately 900% on an initial investment of $100. Our findings not only contribute to academic literature, but have critical practical implications for sports betting. Our study advances the understanding of team sports outcome prediction a burgeoning are in complex system predictions and pave the way for potential profitability and more informed decision making in sports betting markets.Keywords: machine learning, team sports, game outcome prediction, sports betting, profits simulation
Procedia PDF Downloads 93730 Pneumoperitoneum Creation Assisted with Optical Coherence Tomography and Automatic Identification
Authors: Eric Yi-Hsiu Huang, Meng-Chun Kao, Wen-Chuan Kuo
Abstract:
For every laparoscopic surgery, a safe pneumoperitoneumcreation (gaining access to the peritoneal cavity) is the first and essential step. However, closed pneumoperitoneum is usually obtained by blind insertion of a Veress needle into the peritoneal cavity, which may carry potential risks suchas bowel and vascular injury.Until now, there remains no definite measure to visually confirm the position of the needle tip inside the peritoneal cavity. Therefore, this study established an image-guided Veress needle method by combining a fiber probe with optical coherence tomography (OCT). An algorithm was also proposed for determining the exact location of the needle tip through the acquisition of OCT images. Our method not only generates a series of “live” two-dimensional (2D) images during the needle puncture toward the peritoneal cavity but also can eliminate operator variation in image judgment, thus improving peritoneal access safety. This study was approved by the Ethics Committee of Taipei Veterans General Hospital (Taipei VGH IACUC 2020-144). A total of 2400 in vivo OCT images, independent of each other, were acquired from experiments of forty peritoneal punctures on two piglets. Characteristic OCT image patterns could be observed during the puncturing process. The ROC curve demonstrates the discrimination capability of these quantitative image features of the classifier, showing the accuracy of the classifier for determining the inside vs. outside of the peritoneal was 98% (AUC=0.98). In summary, the present study demonstrates the ability of the combination of our proposed automatic identification method and OCT imaging for automatically and objectively identifying the location of the needle tip. OCT images translate the blind closed technique of peritoneal access into a visualized procedure, thus improving peritoneal access safety.Keywords: pneumoperitoneum, optical coherence tomography, automatic identification, veress needle
Procedia PDF Downloads 124729 A Novel Method for Face Detection
Authors: H. Abas Nejad, A. R. Teymoori
Abstract:
Facial expression recognition is one of the open problems in computer vision. Robust neutral face recognition in real time is a major challenge for various supervised learning based facial expression recognition methods. This is due to the fact that supervised methods cannot accommodate all appearance variability across the faces with respect to race, pose, lighting, facial biases, etc. in the limited amount of training data. Moreover, processing each and every frame to classify emotions is not required, as the user stays neutral for the majority of the time in usual applications like video chat or photo album/web browsing. Detecting neutral state at an early stage, thereby bypassing those frames from emotion classification would save the computational power. In this work, we propose a light-weight neutral vs. emotion classification engine, which acts as a preprocessor to the traditional supervised emotion classification approaches. It dynamically learns neutral appearance at Key Emotion (KE) points using a textural statistical model, constructed by a set of reference neutral frames for each user. The proposed method is made robust to various types of user head motions by accounting for affine distortions based on a textural statistical model. Robustness to dynamic shift of KE points is achieved by evaluating the similarities on a subset of neighborhood patches around each KE point using the prior information regarding the directionality of specific facial action units acting on the respective KE point. The proposed method, as a result, improves ER accuracy and simultaneously reduces the computational complexity of ER system, as validated on multiple databases.Keywords: neutral vs. emotion classification, Constrained Local Model, procrustes analysis, Local Binary Pattern Histogram, statistical model
Procedia PDF Downloads 333728 Formation of the Investment Portfolio of Intangible Assets with a Wide Pairwise Comparison Matrix Application
Authors: Gulnara Galeeva
Abstract:
The Analytic Hierarchy Process is widely used in the economic and financial studies, including the formation of investment portfolios. In this study, a generalized method of obtaining a vector of priorities for the case with separate pairwise comparisons of the expert opinion being presented as a set of several equal evaluations on a ratio scale is examined. The author claims that this method allows solving an important and up-to-date problem of excluding vagueness and ambiguity of the expert opinion in the decision making theory. The study describes the authentic wide pairwise comparison matrix. Its application in the formation of the efficient investment portfolio of intangible assets of a small business enterprise with limited funding is considered. The proposed method has been successfully approbated on the practical example of a functioning dental clinic. The result of the study confirms that the wide pairwise comparison matrix can be used as a simple and reliable method for forming the enterprise investment policy. Moreover, a comparison between the method based on the wide pairwise comparison matrix and the classical analytic hierarchy process was conducted. The results of the comparative analysis confirm the correctness of the method based on the wide matrix. The application of a wide pairwise comparison matrix also allows to widely use the statistical methods of experimental data processing for obtaining the vector of priorities. A new method is available for simple users. Its application gives about the same accuracy result as that of the classical hierarchy process. Financial directors of small and medium business enterprises get an opportunity to solve the problem of companies’ investments without resorting to services of analytical agencies specializing in such studies.Keywords: analytic hierarchy process, decision processes, investment portfolio, intangible assets
Procedia PDF Downloads 259727 Towards a Robust Patch Based Multi-View Stereo Technique for Textureless and Occluded 3D Reconstruction
Authors: Ben Haines, Li Bai
Abstract:
Patch based reconstruction methods have been and still are one of the top performing approaches to 3D reconstruction to date. Their local approach to refining the position and orientation of a patch, free of global minimisation and independent of surface smoothness, make patch based methods extremely powerful in recovering fine grained detail of an objects surface. However, patch based approaches still fail to faithfully reconstruct textureless or highly occluded surface regions thus though performing well under lab conditions, deteriorate in industrial or real world situations. They are also computationally expensive. Current patch based methods generate point clouds with holes in texturesless or occluded regions that require expensive energy minimisation techniques to fill and interpolate a high fidelity reconstruction. Such shortcomings hinder the adaptation of the methods for industrial applications where object surfaces are often highly textureless and the speed of reconstruction is an important factor. This paper presents on-going work towards a multi-resolution approach to address the problems, utilizing particle swarm optimisation to reconstruct high fidelity geometry, and increasing robustness to textureless features through an adapted approach to the normalised cross correlation. The work also aims to speed up the reconstruction using advances in GPU technologies and remove the need for costly initialization and expansion. Through the combination of these enhancements, it is the intention of this work to create denser patch clouds even in textureless regions within a reasonable time. Initial results show the potential of such an approach to construct denser point clouds with a comparable accuracy to that of the current top-performing algorithms.Keywords: 3D reconstruction, multiview stereo, particle swarm optimisation, photo consistency
Procedia PDF Downloads 199726 Shifting to Electronic Operative Notes in Plastic surgery
Authors: Samar Mousa, Galini Mavromatidou, Rebecca Shirley
Abstract:
Surgeons carry out numerous operations in the busy burns and plastic surgery department daily. Writing an accurate operation note with all the essential information is crucial for communication not only within the plastics team but also to the multi-disciplinary team looking after the patient, including other specialties, nurses and GPs. The Royal college of surgeons of England, in its guidelines of good surgical practice, mentioned that the surgeon should ensure that there are clear (preferably typed) operative notes for every procedure. The notes should accompany the patient into recovery and to the ward and should give sufficient detail to enable continuity of care by another doctor. The notes should include the Date and time, Elective/emergency procedure, Names of the operating surgeon and assistant, Name of the theatre anesthetist, Operative procedure carried out, Incision, Operative diagnosis, Operative findings, Any problems/complications, Any extra procedure performed and the reason why it was performed, Details of tissue removed, added or altered, Identification of any prosthesis used, including the serial numbers of prostheses and other implanted materials, Details of closure technique, Anticipated blood loss, Antibiotic prophylaxis (where applicable), DVT prophylaxis (where applicable), Detailed postoperative care instructions and Signature. Fourteen random days were chosen in December 2021 to assess the accuracy of operative notes and post-operative care. A total of 163 operative notes were examined. The average completion rates in all domains were 85.4%. An electronic operative note template was designed to cover all domains mentioned in the Royal College of surgeons' good surgical practice. It is kept in the hospital drive for all surgeons to use.Keywords: operative notes, plastic surgery, documentation, electronic
Procedia PDF Downloads 76725 Human-Machine Cooperation in Facial Comparison Based on Likelihood Scores
Authors: Lanchi Xie, Zhihui Li, Zhigang Li, Guiqiang Wang, Lei Xu, Yuwen Yan
Abstract:
Image-based facial features can be classified into category recognition features and individual recognition features. Current automated face recognition systems extract a specific feature vector of different dimensions from a facial image according to their pre-trained neural network. However, to improve the efficiency of parameter calculation, an algorithm generally reduces the image details by pooling. The operation will overlook the details concerned much by forensic experts. In our experiment, we adopted a variety of face recognition algorithms based on deep learning, compared a large number of naturally collected face images with the known data of the same person's frontal ID photos. Downscaling and manual handling were performed on the testing images. The results supported that the facial recognition algorithms based on deep learning detected structural and morphological information and rarely focused on specific markers such as stains and moles. Overall performance, distribution of genuine scores and impostor scores, and likelihood ratios were tested to evaluate the accuracy of biometric systems and forensic experts. Experiments showed that the biometric systems were skilled in distinguishing category features, and forensic experts were better at discovering the individual features of human faces. In the proposed approach, a fusion was performed at the score level. At the specified false accept rate, the framework achieved a lower false reject rate. This paper contributes to improving the interpretability of the objective method of facial comparison and provides a novel method for human-machine collaboration in this field.Keywords: likelihood ratio, automated facial recognition, facial comparison, biometrics
Procedia PDF Downloads 122724 Gnss Aided Photogrammetry for Digital Mapping
Authors: Muhammad Usman Akram
Abstract:
This research work based on GNSS-Aided Photogrammetry for Digital Mapping. It focuses on topographic survey of an area or site which is to be used in future Planning & development (P&D) or can be used for further, examination, exploration, research and inspection. Survey and Mapping in hard-to-access and hazardous areas are very difficult by using traditional techniques and methodologies; as well it is time consuming, labor intensive and has less precision with limited data. In comparison with the advance techniques it is saving with less manpower and provides more precise output with a wide variety of multiple data sets. In this experimentation, Aerial Photogrammetry technique is used where an UAV flies over an area and captures geocoded images and makes a Three-Dimensional Model (3-D Model), UAV operates on a user specified path or area with various parameters; Flight altitude, Ground sampling distance (GSD), Image overlapping, Camera angle etc. For ground controlling, a network of points on the ground would be observed as a Ground Control point (GCP) using Differential Global Positioning System (DGPS) in PPK or RTK mode. Furthermore, that raw data collected by UAV and DGPS will be processed in various Digital image processing programs and Computer Aided Design software. From which as an output we obtain Points Dense Cloud, Digital Elevation Model (DEM) and Ortho-photo. The imagery is converted into geospatial data by digitizing over Ortho-photo, DEM is further converted into Digital Terrain Model (DTM) for contour generation or digital surface. As a result, we get Digital Map of area to be surveyed. In conclusion, we compared processed data with exact measurements taken on site. The error will be accepted if the amount of error is not breached from survey accuracy limits set by concerned institutions.Keywords: photogrammetry, post processing kinematics, real time kinematics, manual data inquiry
Procedia PDF Downloads 12723 A Machine Learning Approach for Detecting and Locating Hardware Trojans
Authors: Kaiwen Zheng, Wanting Zhou, Nan Tang, Lei Li, Yuanhang He
Abstract:
The integrated circuit industry has become a cornerstone of the information society, finding widespread application in areas such as industry, communication, medicine, and aerospace. However, with the increasing complexity of integrated circuits, Hardware Trojans (HTs) implanted by attackers have become a significant threat to their security. In this paper, we proposed a hardware trojan detection method for large-scale circuits. As HTs introduce physical characteristic changes such as structure, area, and power consumption as additional redundant circuits, we proposed a machine-learning-based hardware trojan detection method based on the physical characteristics of gate-level netlists. This method transforms the hardware trojan detection problem into a machine-learning binary classification problem based on physical characteristics, greatly improving detection speed. To address the problem of imbalanced data, where the number of pure circuit samples is far less than that of HTs circuit samples, we used the SMOTETomek algorithm to expand the dataset and further improve the performance of the classifier. We used three machine learning algorithms, K-Nearest Neighbors, Random Forest, and Support Vector Machine, to train and validate benchmark circuits on Trust-Hub, and all achieved good results. In our case studies based on AES encryption circuits provided by trust-hub, the test results showed the effectiveness of the proposed method. To further validate the method’s effectiveness for detecting variant HTs, we designed variant HTs using open-source HTs. The proposed method can guarantee robust detection accuracy in the millisecond level detection time for IC, and FPGA design flows and has good detection performance for library variant HTs.Keywords: hardware trojans, physical properties, machine learning, hardware security
Procedia PDF Downloads 139722 Comparison of an Anthropomorphic PRESAGE® Dosimeter and Radiochromic Film with a Commercial Radiation Treatment Planning System for Breast IMRT: A Feasibility Study
Authors: Khalid Iqbal
Abstract:
This work presents a comparison of an anthropomorphic PRESAGE® dosimeter and radiochromic film measurements with a commercial treatment planning system to determine the feasibility of PRESAGE® for 3D dosimetry in breast IMRT. An anthropomorphic PRESAGE® phantom was created in the shape of a breast phantom. A five-field IMRT plan was generated with a commercially available treatment planning system and delivered to the PRESAGE® phantom. The anthropomorphic PRESAGE® was scanned with the Duke midsized optical CT scanner (DMOS-RPC) and the OD distribution was converted to dose. Comparisons were performed between the dose distribution calculated with the Pinnacle3 treatment planning system, PRESAGE®, and EBT2 film measurements. DVHs, gamma maps, and line profiles were used to evaluate the agreement. Gamma map comparisons showed that Pinnacle3 agreed with PRESAGE® as greater than 95% of comparison points for the PTV passed a ± 3%/± 3 mm criterion when the outer 8 mm of phantom data were discluded. Edge artifacts were observed in the optical CT reconstruction, from the surface to approximately 8 mm depth. These artifacts resulted in dose differences between Pinnacle3 and PRESAGE® of up to 5% between the surface and a depth of 8 mm and decreased with increasing depth in the phantom. Line profile comparisons between all three independent measurements yielded a maximum difference of 2% within the central 80% of the field width. For the breast IMRT plan studied, the Pinnacle3 calculations agreed with PRESAGE® measurements to within the ±3%/± 3 mm gamma criterion. This work demonstrates the feasibility of the PRESAGE® to be fashioned into anthropomorphic shape, and establishes the accuracy of Pinnacle3 for breast IMRT. Furthermore, these data have established the groundwork for future investigations into 3D dosimetry with more complex anthropomorphic phantoms.Keywords: 3D dosimetry, PRESAGE®, IMRT, QA, EBT2 GAFCHROMIC film
Procedia PDF Downloads 403721 Evaluating the Dosimetric Performance for 3D Treatment Planning System for Wedged and Off-Axis Fields
Authors: Nashaat A. Deiab, Aida Radwan, Mohamed S. Yahiya, Mohamed Elnagdy, Rasha Moustafa
Abstract:
This study is to evaluate the dosimetric performance of our institution's 3D treatment planning system for wedged and off-axis 6MV photon beams, guided by the recommended QA tests documented in the AAPM TG53; NCS report 15 test packages, IAEA TRS 430 and ESTRO booklet no.7. The study was performed for Elekta Precise linear accelerator designed for clinical range of 4, 6 and 15 MV photon beams with asymmetric jaws and fully integrated multileaf collimator that enables high conformance to target with sharp field edges. Ten tests were applied on solid water equivalent phantom along with 2D array dose detection system. The calculated doses using 3D treatment planning system PrecisePLAN were compared with measured doses to make sure that the dose calculations are accurate for simple situations such as square and elongated fields, different SSD, beam modifiers e.g. wedges, blocks, MLC-shaped fields and asymmetric collimator settings. The QA results showed dosimetric accuracy of the TPS within the specified tolerance limits. Except for large elongated wedged field, the central axis and outside central axis have errors of 0.2% and 0.5%, respectively, and off- planned and off-axis elongated fields the region outside the central axis of the beam errors are 0.2% and 1.1%, respectively. The dosimetric investigated results yielded differences within the accepted tolerance level as recommended. Differences between dose values predicted by the TPS and measured values at the same point are the result from limitations of the dose calculation, uncertainties in the measurement procedure, or fluctuations in the output of the accelerator.Keywords: quality assurance, dose calculation, wedged fields, off-axis fields, 3D treatment planning system, photon beam
Procedia PDF Downloads 434720 High-Resolution Flood Hazard Mapping Using Two-Dimensional Hydrodynamic Model Anuga: Case Study of Jakarta, Indonesia
Authors: Hengki Eko Putra, Dennish Ari Putro, Tri Wahyu Hadi, Edi Riawan, Junnaedhi Dewa Gede, Aditia Rojali, Fariza Dian Prasetyo, Yudhistira Satya Pribadi, Dita Fatria Andarini, Mila Khaerunisa, Raditya Hanung Prakoswa
Abstract:
Catastrophe risk management can only be done if we are able to calculate the exposed risks. Jakarta is an important city economically, socially, and politically and in the same time exposed to severe floods. On the other hand, flood risk calculation is still very limited in the area. This study has calculated the risk of flooding for Jakarta using 2-Dimensional Model ANUGA. 2-Dimensional model ANUGA and 1-Dimensional Model HEC-RAS are used to calculate the risk of flooding from 13 major rivers in Jakarta. ANUGA can simulate physical and dynamical processes between the streamflow against river geometry and land cover to produce a 1-meter resolution inundation map. The value of streamflow as an input for the model obtained from hydrological analysis on rainfall data using hydrologic model HEC-HMS. The probabilistic streamflow derived from probabilistic rainfall using statistical distribution Log-Pearson III, Normal and Gumbel, through compatibility test using Chi Square and Smirnov-Kolmogorov. Flood event on 2007 is used as a comparison to evaluate the accuracy of model output. Property damage estimations were calculated based on flood depth for 1, 5, 10, 25, 50, and 100 years return period against housing value data from the BPS-Statistics Indonesia, Centre for Research and Development of Housing and Settlements, Ministry of Public Work Indonesia. The vulnerability factor was derived from flood insurance claim. Jakarta's flood loss estimation for the return period of 1, 5, 10, 25, 50, and 100 years, respectively are Rp 1.30 t; Rp 16.18 t; Rp 16.85 t; Rp 21.21 t; Rp 24.32 t; and Rp 24.67 t of the total value of building Rp 434.43 t.Keywords: 2D hydrodynamic model, ANUGA, flood, flood modeling
Procedia PDF Downloads 269719 Feature Evaluation Based on Random Subspace and Multiple-K Ensemble
Authors: Jaehong Yu, Seoung Bum Kim
Abstract:
Clustering analysis can facilitate the extraction of intrinsic patterns in a dataset and reveal its natural groupings without requiring class information. For effective clustering analysis in high dimensional datasets, unsupervised dimensionality reduction is an important task. Unsupervised dimensionality reduction can generally be achieved by feature extraction or feature selection. In many situations, feature selection methods are more appropriate than feature extraction methods because of their clear interpretation with respect to the original features. The unsupervised feature selection can be categorized as feature subset selection and feature ranking method, and we focused on unsupervised feature ranking methods which evaluate the features based on their importance scores. Recently, several unsupervised feature ranking methods were developed based on ensemble approaches to achieve their higher accuracy and stability. However, most of the ensemble-based feature ranking methods require the true number of clusters. Furthermore, these algorithms evaluate the feature importance depending on the ensemble clustering solution, and they produce undesirable evaluation results if the clustering solutions are inaccurate. To address these limitations, we proposed an ensemble-based feature ranking method with random subspace and multiple-k ensemble (FRRM). The proposed FRRM algorithm evaluates the importance of each feature with the random subspace ensemble, and all evaluation results are combined with the ensemble importance scores. Moreover, FRRM does not require the determination of the true number of clusters in advance through the use of the multiple-k ensemble idea. Experiments on various benchmark datasets were conducted to examine the properties of the proposed FRRM algorithm and to compare its performance with that of existing feature ranking methods. The experimental results demonstrated that the proposed FRRM outperformed the competitors.Keywords: clustering analysis, multiple-k ensemble, random subspace-based feature evaluation, unsupervised feature ranking
Procedia PDF Downloads 329718 An Unsupervised Domain-Knowledge Discovery Framework for Fake News Detection
Authors: Yulan Wu
Abstract:
With the rapid development of social media, the issue of fake news has gained considerable prominence, drawing the attention of both the public and governments. The widespread dissemination of false information poses a tangible threat across multiple domains of society, including politics, economy, and health. However, much research has concentrated on supervised training models within specific domains, their effectiveness diminishes when applied to identify fake news across multiple domains. To solve this problem, some approaches based on domain labels have been proposed. By segmenting news to their specific area in advance, judges in the corresponding field may be more accurate on fake news. However, these approaches disregard the fact that news records can pertain to multiple domains, resulting in a significant loss of valuable information. In addition, the datasets used for training must all be domain-labeled, which creates unnecessary complexity. To solve these problems, an unsupervised domain knowledge discovery framework for fake news detection is proposed. Firstly, to effectively retain the multidomain knowledge of the text, a low-dimensional vector for each news text to capture domain embeddings is generated. Subsequently, a feature extraction module utilizing the unsupervisedly discovered domain embeddings is used to extract the comprehensive features of news. Finally, a classifier is employed to determine the authenticity of the news. To verify the proposed framework, a test is conducted on the existing widely used datasets, and the experimental results demonstrate that this method is able to improve the detection performance for fake news across multiple domains. Moreover, even in datasets that lack domain labels, this method can still effectively transfer domain knowledge, which can educe the time consumed by tagging without sacrificing the detection accuracy.Keywords: fake news, deep learning, natural language processing, multiple domains
Procedia PDF Downloads 82717 Hyper Parameter Optimization of Deep Convolutional Neural Networks for Pavement Distress Classification
Authors: Oumaima Khlifati, Khadija Baba
Abstract:
Pavement distress is the main factor responsible for the deterioration of road structure durability, damage vehicles, and driver comfort. Transportation agencies spend a high proportion of their funds on pavement monitoring and maintenance. The auscultation of pavement distress was based on the manual survey, which was extremely time consuming, labor intensive, and required domain expertise. Therefore, the automatic distress detection is needed to reduce the cost of manual inspection and avoid more serious damage by implementing the appropriate remediation actions at the right time. Inspired by recent deep learning applications, this paper proposes an algorithm for automatic road distress detection and classification using on the Deep Convolutional Neural Network (DCNN). In this study, the types of pavement distress are classified as transverse or longitudinal cracking, alligator, pothole, and intact pavement. The dataset used in this work is composed of public asphalt pavement images. In order to learn the structure of the different type of distress, the DCNN models are trained and tested as a multi-label classification task. In addition, to get the highest accuracy for our model, we adjust the structural optimization hyper parameters such as the number of convolutions and max pooling, filers, size of filters, loss functions, activation functions, and optimizer and fine-tuning hyper parameters that conclude batch size and learning rate. The optimization of the model is executed by checking all feasible combinations and selecting the best performing one. The model, after being optimized, performance metrics is calculated, which describe the training and validation accuracies, precision, recall, and F1 score.Keywords: distress pavement, hyperparameters, automatic classification, deep learning
Procedia PDF Downloads 81716 Creating Database and Building 3D Geological Models: A Case Study on Bac Ai Pumped Storage Hydropower Project
Authors: Nguyen Chi Quang, Nguyen Duong Tri Nguyen
Abstract:
This article is the first step to research and outline the structure of the geotechnical database in the geological survey of a power project; in the context of this report creating the database that has been carried out for the Bac Ai pumped storage hydropower project. For the purpose of providing a method of organizing and storing geological and topographic survey data and experimental results in a spatial database, the RockWorks software is used to bring optimal efficiency in the process of exploiting, using, and analyzing data in service of the design work in the power engineering consulting. Three-dimensional (3D) geotechnical models are created from the survey data: such as stratigraphy, lithology, porosity, etc. The results of the 3D geotechnical model in the case of Bac Ai pumped storage hydropower project include six closely stacked stratigraphic formations by Horizons method, whereas modeling of engineering geological parameters is performed by geostatistical methods. The accuracy and reliability assessments are tested through error statistics, empirical evaluation, and expert methods. The three-dimensional model analysis allows better visualization of volumetric calculations, excavation and backfilling of the lake area, tunneling of power pipelines, and calculation of on-site construction material reserves. In general, the application of engineering geological modeling makes the design work more intuitive and comprehensive, helping construction designers better identify and offer the most optimal design solutions for the project. The database always ensures the update and synchronization, as well as enables 3D modeling of geological and topographic data to integrate with the designed data according to the building information modeling. This is also the base platform for BIM & GIS integration.Keywords: database, engineering geology, 3D Model, RockWorks, Bac Ai pumped storage hydropower project
Procedia PDF Downloads 156715 Developing a DNN Model for the Production of Biogas From a Hybrid BO-TPE System in an Anaerobic Wastewater Treatment Plant
Authors: Hadjer Sadoune, Liza Lamini, Scherazade Krim, Amel Djouadi, Rachida Rihani
Abstract:
Deep neural networks are highly regarded for their accuracy in predicting intricate fermentation processes. Their ability to learn from a large amount of datasets through artificial intelligence makes them particularly effective models. The primary obstacle in improving the performance of these models is to carefully choose the suitable hyperparameters, including the neural network architecture (number of hidden layers and hidden units), activation function, optimizer, learning rate, and other relevant factors. This study predicts biogas production from real wastewater treatment plant data using a sophisticated approach: hybrid Bayesian optimization with a tree-structured Parzen estimator (BO-TPE) for an optimised deep neural network (DNN) model. The plant utilizes an Upflow Anaerobic Sludge Blanket (UASB) digester that treats industrial wastewater from soft drinks and breweries. The digester has a working volume of 1574 m3 and a total volume of 1914 m3. Its internal diameter and height were 19 and 7.14 m, respectively. The data preprocessing was conducted with meticulous attention to preserving data quality while avoiding data reduction. Three normalization techniques were applied to the pre-processed data (MinMaxScaler, RobustScaler and StandardScaler) and compared with the Non-Normalized data. The RobustScaler approach has strong predictive ability for estimating the volume of biogas produced. The highest predicted biogas volume was 2236.105 Nm³/d, with coefficient of determination (R2), mean absolute error (MAE), and root mean square error (RMSE) values of 0.712, 164.610, and 223.429, respectively.Keywords: anaerobic digestion, biogas production, deep neural network, hybrid bo-tpe, hyperparameters tuning
Procedia PDF Downloads 29714 Comparison of Support Vector Machines and Artificial Neural Network Classifiers in Characterizing Threatened Tree Species Using Eight Bands of WorldView-2 Imagery in Dukuduku Landscape, South Africa
Authors: Galal Omer, Onisimo Mutanga, Elfatih M. Abdel-Rahman, Elhadi Adam
Abstract:
Threatened tree species (TTS) play a significant role in ecosystem functioning and services, land use dynamics, and other socio-economic aspects. Such aspects include ecological, economic, livelihood, security-based, and well-being benefits. The development of techniques for mapping and monitoring TTS is thus critical for understanding the functioning of ecosystems. The advent of advanced imaging systems and supervised learning algorithms has provided an opportunity to classify TTS over fragmenting landscape. Recently, vegetation maps have been produced using advanced imaging systems such as WorldView-2 (WV-2) and robust classification algorithms such as support vectors machines (SVM) and artificial neural network (ANN). However, delineation of TTS in a fragmenting landscape using high resolution imagery has widely remained elusive due to the complexity of the species structure and their distribution. Therefore, the objective of the current study was to examine the utility of the advanced WV-2 data for mapping TTS in the fragmenting Dukuduku indigenous forest of South Africa using SVM and ANN classification algorithms. The results showed the robustness of the two machine learning algorithms with an overall accuracy (OA) of 77.00% (total disagreement = 23.00%) for SVM and 75.00% (total disagreement = 25.00%) for ANN using all eight bands of WV-2 (8B). This study concludes that SVM and ANN classification algorithms with WV-2 8B have the potential to classify TTS in the Dukuduku indigenous forest. This study offers relatively accurate information that is important for forest managers to make informed decisions regarding management and conservation protocols of TTS.Keywords: artificial neural network, threatened tree species, indigenous forest, support vector machines
Procedia PDF Downloads 508713 Evaluation and Analysis of ZigBee-Based Wireless Sensor Network: Home Monitoring as Case Study
Authors: Omojokun G. Aju, Adedayo O. Sule
Abstract:
ZigBee wireless sensor and control network is one of the most popularly deployed wireless technologies in recent years. This is because ZigBee is an open standard lightweight, low-cost, low-speed, low-power protocol that allows true operability between systems. It is built on existing IEEE 802.15.4 protocol and therefore combines the IEEE 802.15.4 features and newly added features to meet required functionalities thereby finding applications in wide variety of wireless networked systems. ZigBee‘s current focus is on embedded applications of general-purpose, inexpensive, self-organising networks which requires low to medium data rates, high number of nodes and very low power consumption such as home/industrial automation, embedded sensing, medical data collection, smart lighting, safety and security sensor networks, and monitoring systems. Although the ZigBee design specification includes security features to protect data communication confidentiality and integrity, however, when simplicity and low-cost are the goals, security is normally traded-off. A lot of researches have been carried out on ZigBee technology in which emphasis has mainly been placed on ZigBee network performance characteristics such as energy efficiency, throughput, robustness, packet delay and delivery ratio in different scenarios and applications. This paper investigate and analyse the data accuracy, network implementation difficulties and security challenges of ZigBee network applications in star-based and mesh-based topologies with emphases on its home monitoring application using the ZigBee ProBee ZE-10 development boards for the network setup. The paper also expose some factors that need to be considered when designing ZigBee network applications and suggest ways in which ZigBee network can be designed to provide more resilient to network attacks.Keywords: home monitoring, IEEE 802.14.5, topology, wireless security, wireless sensor network (WSN), ZigBee
Procedia PDF Downloads 373712 A Damage Level Assessment Model for Extra High Voltage Transmission Towers
Authors: Huan-Chieh Chiu, Hung-Shuo Wu, Chien-Hao Wang, Yu-Cheng Yang, Ching-Ya Tseng, Joe-Air Jiang
Abstract:
Power failure resulting from tower collapse due to violent seismic events might bring enormous and inestimable losses. The Chi-Chi earthquake, for example, strongly struck Taiwan and caused huge damage to the power system on September 21, 1999. Nearly 10% of extra high voltage (EHV) transmission towers were damaged in the earthquake. Therefore, seismic hazards of EHV transmission towers should be monitored and evaluated. The ultimate goal of this study is to establish a damage level assessment model for EHV transmission towers. The data of earthquakes provided by Taiwan Central Weather Bureau serve as a reference and then lay the foundation for earthquake simulations and analyses afterward. Some parameters related to the damage level of each point of an EHV tower are simulated and analyzed by the data from monitoring stations once an earthquake occurs. Through the Fourier transform, the seismic wave is then analyzed and transformed into different wave frequencies, and the data would be shown through a response spectrum. With this method, the seismic frequency which damages EHV towers the most is clearly identified. An estimation model is built to determine the damage level caused by a future seismic event. Finally, instead of relying on visual observation done by inspectors, the proposed model can provide a power company with the damage information of a transmission tower. Using the model, manpower required by visual observation can be reduced, and the accuracy of the damage level estimation can be substantially improved. Such a model is greatly useful for health and construction monitoring because of the advantages of long-term evaluation of structural characteristics and long-term damage detection.Keywords: damage level monitoring, drift ratio, fragility curve, smart grid, transmission tower
Procedia PDF Downloads 296711 Alternating Expectation-Maximization Algorithm for a Bilinear Model in Isoform Quantification from RNA-Seq Data
Authors: Wenjiang Deng, Tian Mou, Yudi Pawitan, Trung Nghia Vu
Abstract:
Estimation of isoform-level gene expression from RNA-seq data depends on simplifying assumptions, such as uniform reads distribution, that are easily violated in real data. Such violations typically lead to biased estimates. Most existing methods provide a bias correction step(s), which is based on biological considerations, such as GC content–and applied in single samples separately. The main problem is that not all biases are known. For example, new technologies such as single-cell RNA-seq (scRNA-seq) may introduce new sources of bias not seen in bulk-cell data. This study introduces a method called XAEM based on a more flexible and robust statistical model. Existing methods are essentially based on a linear model Xβ, where the design matrix X is known and derived based on the simplifying assumptions. In contrast, XAEM considers Xβ as a bilinear model with both X and β unknown. Joint estimation of X and β is made possible by simultaneous analysis of multi-sample RNA-seq data. Compared to existing methods, XAEM automatically performs empirical correction of potentially unknown biases. XAEM implements an alternating expectation-maximization (AEM) algorithm, alternating between estimation of X and β. For speed XAEM utilizes quasi-mapping for read alignment, thus leading to a fast algorithm. Overall XAEM performs favorably compared to other recent advanced methods. For simulated datasets, XAEM obtains higher accuracy for multiple-isoform genes, particularly for paralogs. In a differential-expression analysis of a real scRNA-seq dataset, XAEM achieves substantially greater rediscovery rates in an independent validation set.Keywords: alternating EM algorithm, bias correction, bilinear model, gene expression, RNA-seq
Procedia PDF Downloads 135710 Loss Function Optimization for CNN-Based Fingerprint Anti-Spoofing
Authors: Yehjune Heo
Abstract:
As biometric systems become widely deployed, the security of identification systems can be easily attacked by various spoof materials. This paper contributes to finding a reliable and practical anti-spoofing method using Convolutional Neural Networks (CNNs) based on the types of loss functions and optimizers. The types of CNNs used in this paper include AlexNet, VGGNet, and ResNet. By using various loss functions including Cross-Entropy, Center Loss, Cosine Proximity, and Hinge Loss, and various loss optimizers which include Adam, SGD, RMSProp, Adadelta, Adagrad, and Nadam, we obtained significant performance changes. We realize that choosing the correct loss function for each model is crucial since different loss functions lead to different errors on the same evaluation. By using a subset of the Livdet 2017 database, we validate our approach to compare the generalization power. It is important to note that we use a subset of LiveDet and the database is the same across all training and testing for each model. This way, we can compare the performance, in terms of generalization, for the unseen data across all different models. The best CNN (AlexNet) with the appropriate loss function and optimizers result in more than 3% of performance gain over the other CNN models with the default loss function and optimizer. In addition to the highest generalization performance, this paper also contains the models with high accuracy associated with parameters and mean average error rates to find the model that consumes the least memory and computation time for training and testing. Although AlexNet has less complexity over other CNN models, it is proven to be very efficient. For practical anti-spoofing systems, the deployed version should use a small amount of memory and should run very fast with high anti-spoofing performance. For our deployed version on smartphones, additional processing steps, such as quantization and pruning algorithms, have been applied in our final model.Keywords: anti-spoofing, CNN, fingerprint recognition, loss function, optimizer
Procedia PDF Downloads 131709 The Utility of Sonographic Features of Lymph Nodes during EBUS-TBNA for Predicting Malignancy
Authors: Atefeh Abedini, Fatemeh Razavi, Mihan Pourabdollah Toutkaboni, Hossein Mehravaran, Arda Kiani
Abstract:
In countries with the highest prevalence of tuberculosis, such as Iran, the differentiation of malignant tumors from non-malignant is very important. In this study, which was conducted for the first time among the Iranian population, the utility of the ultrasonographic morphological characteristics in patients undergoing EBUS was used to distinguish the non-malignant versus malignant lymph nodes. The morphological characteristics of lymph nodes, which consist of size, shape, vascular pattern, echogenicity, margin, coagulation necrosis sign, calcification, and central hilar structure, were obtained during Endobronchial Ultrasound-Guided Trans-Bronchial Needle Aspiration and were compared with the final pathology results. During this study period, a total of 253 lymph nodes were evaluated in 93 cases. Round shape, non-hilar vascular pattern, heterogeneous echogenicity, hyperechogenicity, distinct margin, and the presence of necrosis sign were significantly higher in malignant nodes. On the other hand, the presence of calcification and also central hilar structure were significantly higher in the benign nodes (p-value ˂ 0.05). Multivariate logistic regression showed that size>1 cm, heterogeneous echogenicity, hyperechogenicity, the presence of necrosis signs and, the absence of central hilar structure are independent predictive factors for malignancy. The accuracy of each of the aforementioned factors is 42.29 %, 71.54 %, 71.90 %, 73.51 %, and 65.61 %, respectively. Of 74 malignant lymph nodes, 100% had at least one of these independent factors. According to our results, the morphological characteristics of lymph nodes based on Endobronchial Ultrasound-Guided Trans-Bronchial Needle Aspiration can play a role in the prediction of malignancy.Keywords: EBUS-TBNA, malignancy, nodal characteristics, pathology
Procedia PDF Downloads 129708 The Use of Unmanned Aerial System (UAS) in Improving the Measurement System on the Example of Textile Heaps
Authors: Arkadiusz Zurek
Abstract:
The potential of using drones is visible in many areas of logistics, especially in terms of their use for monitoring and control of many processes. The technologies implemented in the last decade concern new possibilities for companies that until now have not even considered them, such as warehouse inventories. Unmanned aerial vehicles are no longer seen as a revolutionary tool for Industry 4.0, but rather as tools in the daily work of factories and logistics operators. The research problem is to develop a method for measuring the weight of goods in a selected link of the clothing supply chain by drones. However, the purpose of this article is to analyze the causes of errors in traditional measurements, and then to identify adverse events related to the use of drones for the inventory of a heap of textiles intended for production purposes. On this basis, it will be possible to develop guidelines to eliminate the causes of these events in the measurement process using drones. In a real environment, work was carried out to determine the volume and weight of textiles, including, among others, weighing a textile sample to determine the average density of the assortment, establishing a local geodetic network, terrestrial laser scanning and photogrammetric raid using an unmanned aerial vehicle. As a result of the analysis of measurement data obtained in the facility, the volume and weight of the assortment and the accuracy of their determination were determined. In this article, this work presents how such heaps are currently being tested, what adverse events occur, indicate and describes the current use of photogrammetric techniques of this type of measurements so far performed by external drones for the inventory of wind farms or construction of the station and compare them with the measurement system of the aforementioned textile heap inside a large-format facility.Keywords: drones, unmanned aerial system, UAS, indoor system, security, process automation, cost optimization, photogrammetry, risk elimination, industry 4.0
Procedia PDF Downloads 76707 Efficacy of Phonological Awareness Intervention for People with Language Impairment
Authors: I. Wardana Ketut, I. Suparwa Nyoman
Abstract:
This study investigated the form and characteristic of speech sound produced by three Balinese subjects who have recovered from aphasia as well as intervened their language impairment on side of linguistic and neuronal aspects of views. The failure of judging the speech sound was caused by impairment of motor cortex that indicated there were lesions in left hemispheric language zone. Sound articulation phenomena were in the forms of phonemes deletion, replacement or assimilation in individual words and meaning building for anomic aphasia. Therefore, the Balinese sound patterns were stimulated by showing pictures to the subjects and recorded to recognize what individual consonants or vowels they unclearly produced and to find out how the sound disorder occurred. The physiology of sound production by subject’s speech organs could not only show the accuracy of articulation but also any level of severity the lesion they suffered from. The subjects’ speech sounds were investigated, classified and analyzed to know how poor the lingual units were and observed to clarify weaknesses of sound characters occurred either for place or manner of articulation. Many fricative and stopped consonants were replaced by glottal or palatal sounds because the cranial nerve, such as facial, trigeminal, and hypoglossal underwent impairment after the stroke. The phonological intervention was applied through a technique called phonemic articulation drill and the examination was conducted to know any change has been obtained. The finding informed that some weak articulation turned into clearer sound and simple meaning of language has been conveyed. The hierarchy of functional parts of brain played important role of language formulation and processing. From this finding, it can be clearly emphasized that this study supports the role of right hemisphere in recovery from aphasia is associated with functional brain reorganization.Keywords: aphasia, intervention, phonology, stroke
Procedia PDF Downloads 190706 Reliability Analysis of Construction Schedule Plan Based on Building Information Modelling
Authors: Lu Ren, You-Liang Fang, Yan-Gang Zhao
Abstract:
In recent years, the application of BIM (Building Information Modelling) to construction schedule plan has been the focus of more and more researchers. In order to assess the reasonable level of the BIM-based construction schedule plan, that is whether the schedule can be completed on time, some researchers have introduced reliability theory to evaluate. In the process of evaluation, the uncertain factors affecting the construction schedule plan are regarded as random variables, and probability distributions of the random variables are assumed to be normal distribution, which is determined using two parameters evaluated from the mean and standard deviation of statistical data. However, in practical engineering, most of the uncertain influence factors are not normal random variables. So the evaluation results of the construction schedule plan will be unreasonable under the assumption that probability distributions of random variables submitted to the normal distribution. Therefore, in order to get a more reasonable evaluation result, it is necessary to describe the distribution of random variables more comprehensively. For this purpose, cubic normal distribution is introduced in this paper to describe the distribution of arbitrary random variables, which is determined by the first four moments (mean, standard deviation, skewness and kurtosis). In this paper, building the BIM model firstly according to the design messages of the structure and making the construction schedule plan based on BIM, then the cubic normal distribution is used to describe the distribution of the random variables due to the collecting statistical data of the random factors influencing construction schedule plan. Next the reliability analysis of the construction schedule plan based on BIM can be carried out more reasonably. Finally, the more accurate evaluation results can be given providing reference for the implementation of the actual construction schedule plan. In the last part of this paper, the more efficiency and accuracy of the proposed methodology for the reliability analysis of the construction schedule plan based on BIM are conducted through practical engineering case.Keywords: BIM, construction schedule plan, cubic normal distribution, reliability analysis
Procedia PDF Downloads 135705 Evaluation of Automated Analyzers of Polycyclic Aromatic Hydrocarbons and Black Carbon in a Coke Oven Plant by Comparison with Analytical Methods
Authors: L. Angiuli, L. Trizio, R. Giua, A. Digilio, M. Tutino, P. Dambruoso, F. Mazzone, C. M. Placentino
Abstract:
In the winter of 2014 a series of measurements were performed to evaluate the behavior of real-time PAHs and black carbon analyzers in a coke oven plant located in Taranto, a city of Southern Italy. Data were collected both insides than outside the plant, at air quality monitoring sites. Contemporary measures of PM2.5 and PM1 were performed. Particle-bound PAHs were measured by two methods: (1) aerosol photoionization using an Ecochem PAS 2000 analyzer, (2) PM2.5 and PM1 quartz filter collection and analysis by gas chromatography/mass spectrometry (GC/MS). Black carbon was determined both in real-time by Magee Aethalometer AE22 analyzer than by semi-continuous Sunset Lab EC/OC instrument. Detected PM2.5 and PM1 levels were higher inside than outside the plant while PAHs real-time values were higher outside than inside. As regards PAHs, inside the plant Ecochem PAS 2000 revealed concentrations not significantly different from those determined on the filter during low polluted days, but at increasing concentrations the automated instrument underestimated PAHs levels. At the external site, Ecochem PAS 2000 real-time concentrations were steadily higher than those on the filter. In the same way, real-time black carbon values were constantly lower than EC concentrations obtained by Sunset EC/OC in the inner site, while outside the plant real-time values were comparable to Sunset EC values. Results showed that in a coke plant real-time analyzers of PAHs and black carbon in the factory configuration provide qualitative information, with no accuracy and leading to the underestimation of the concentration. A site specific calibration is needed for these instruments before their installation in high polluted sites.Keywords: black carbon, coke oven plant, PAH, PAS, aethalometer
Procedia PDF Downloads 339704 Evaluating Structural Crack Propagation Induced by Soundless Chemical Demolition Agent Using an Energy Release Rate Approach
Authors: Shyaka Eugene
Abstract:
The efficient and safe demolition of structures is a critical challenge in civil engineering and construction. This study focuses on the development of optimal demolition strategies by investigating the crack propagation behavior in beams induced by soundless cracking agents. It is commonly used in controlled demolition and has gained prominence due to its non-explosive and environmentally friendly nature. This research employs a comprehensive experimental and computational approach to analyze the crack initiation, propagation, and eventual failure in beams subjected to soundless cracking agents. Experimental testing involves the application of various cracking agents under controlled conditions to understand their effects on the structural integrity of beams. High-resolution imaging and strain measurements are used to capture the crack propagation process. In parallel, numerical simulations are conducted using advanced finite element analysis (FEA) techniques to model crack propagation in beams, considering various parameters such as cracking agent composition, loading conditions, and beam properties. The FEA models are validated against experimental results, ensuring their accuracy in predicting crack propagation patterns. The findings of this study provide valuable insights into optimizing demolition strategies, allowing engineers and demolition experts to make informed decisions regarding the selection of cracking agents, their application techniques, and structural reinforcement methods. Ultimately, this research contributes to enhancing the safety, efficiency, and sustainability of demolition practices in the construction industry, reducing environmental impact and ensuring the protection of adjacent structures and the surrounding environment.Keywords: expansion pressure, energy release rate, soundless chemical demolition agent, crack propagation
Procedia PDF Downloads 53703 Finite Element Analysis for Earing Prediction Incorporating the BBC2003 Material Model with Fully Implicit Integration Method: Derivation and Numerical Algorithm
Authors: Sajjad Izadpanah, Seyed Hadi Ghaderi, Morteza Sayah Irani, Mahdi Gerdooei
Abstract:
In this research work, a sophisticated yield criterion known as BBC2003, capable of describing planar anisotropic behaviors of aluminum alloy sheets, was integrated into the commercial finite element code ABAQUS/Standard via a user subroutine. The complete formulation of the implementation process using a fully implicit integration scheme, i.e., the classic backward Euler method, is presented, and relevant aspects of the yield criterion are introduced. In order to solve nonlinear differential and algebraic equations, the line-search algorithm was adopted in the user-defined material subroutine (UMAT) to expand the convergence domain of the iterative Newton-Raphson method. The developed subroutine was used to simulate a challenging computational problem with complex stress states, i.e., deep drawing of an anisotropic aluminum alloy AA3105. The accuracy and stability of the developed subroutine were confirmed by comparing the numerically predicted earing and thickness variation profiles with the experimental results, which showed an excellent agreement between numerical and experimental earing and thickness profiles. The integration of the BBC2003 yield criterion into ABAQUS/Standard represents a significant contribution to the field of computational mechanics and provides a useful tool for analyzing the mechanical behavior of anisotropic materials subjected to complex loading conditions.Keywords: BBC2003 yield function, plastic anisotropy, fully implicit integration scheme, line search algorithm, explicit and implicit integration schemes
Procedia PDF Downloads 67