Search results for: extreme leaning machine (ELM)
3212 Customer Churn Prediction by Using Four Machine Learning Algorithms Integrating Features Selection and Normalization in the Telecom Sector
Authors: Alanoud Moraya Aldalan, Abdulaziz Almaleh
Abstract:
A crucial component of maintaining a customer-oriented business as in the telecom industry is understanding the reasons and factors that lead to customer churn. Competition between telecom companies has greatly increased in recent years. It has become more important to understand customers’ needs in this strong market of telecom industries, especially for those who are looking to turn over their service providers. So, predictive churn is now a mandatory requirement for retaining those customers. Machine learning can be utilized to accomplish this. Churn Prediction has become a very important topic in terms of machine learning classification in the telecommunications industry. Understanding the factors of customer churn and how they behave is very important to building an effective churn prediction model. This paper aims to predict churn and identify factors of customers’ churn based on their past service usage history. Aiming at this objective, the study makes use of feature selection, normalization, and feature engineering. Then, this study compared the performance of four different machine learning algorithms on the Orange dataset: Logistic Regression, Random Forest, Decision Tree, and Gradient Boosting. Evaluation of the performance was conducted by using the F1 score and ROC-AUC. Comparing the results of this study with existing models has proven to produce better results. The results showed the Gradients Boosting with feature selection technique outperformed in this study by achieving a 99% F1-score and 99% AUC, and all other experiments achieved good results as well.Keywords: machine learning, gradient boosting, logistic regression, churn, random forest, decision tree, ROC, AUC, F1-score
Procedia PDF Downloads 1283211 The Rise of Blue Water Navy and its Implication for the Region
Authors: Riddhi Chopra
Abstract:
Alfred Thayer Mahan described the sea as a ‘great common,’ which would serve as a medium for communication, trade, and transport. The seas of Asia are witnessing an intriguing historical anomaly – rise of an indigenous maritime power against the backdrop of US domination over the region. As China transforms from an inward leaning economy to an outward-leaning economy, it has become increasingly dependent on the global sea; as a result, we witness an evolution in its maritime strategy from near seas defense to far seas deployment strategies. It is not only patrolling the international waters but has also built a network of civilian and military infrastructure across the disputed oceanic expanse. The paper analyses the reorientation of China from a naval power to a blue water navy in an era of extensive globalisation. The actions of the Chinese have created a zone of high alert amongst its neighbors such as Japan, Philippines, Vietnam and North Korea. These nations are trying to align themselves so as to counter China’s growing brinkmanship, but China has been pursuing claims through a carefully calibrated strategy in the region shunning any coercive measures taken by other forces. If China continues to expand its maritime boundaries, its neighbors – all smaller and weaker Asian nations would be limited to a narrow band of the sea along its coastlines. Hence it is essential for the US to intervene and support its allies to offset Chinese supremacy. The paper intends to provide a profound analysis over the disputes in South China Sea and East China Sea focusing on Philippines and Japan respectively. Moreover, the paper attempts to give an account of US involvement in the region and its alignment with its South Asian allies. The geographic dynamics is said the breed a national coalition dominating the strategic ambitions of China as well as the weak littoral states. China has conducted behind the scenes diplomacy trying to persuade its neighbors to support its position on the territorial disputes. These efforts have been successful in creating fault lines in ASEAN thereby undermining regional integrity to reach a consensus on the issue. Chinese diplomatic efforts have also forced the US to revisit its foreign policy and engage with players like Cambodia and Laos. The current scenario in the SCS points to a strong Chinese hold trying to outspace all others with no regards to International law. Chinese activities are in contrast with US principles like Freedom of Navigation thereby signaling US to take bold actions to prevent Chinese hegemony in the region. The paper ultimately seeks to explore the changing power dynamics among various claimants where a rival superpower like US can pursue the traditional policy of alliance formation play a decisive role in changing the status quo in the arena, consequently determining the future trajectory.Keywords: China, East China Sea, South China Sea, USA
Procedia PDF Downloads 2373210 Investigation of Extreme Gradient Boosting Model Prediction of Soil Strain-Shear Modulus
Authors: Ehsan Mehryaar, Reza Bushehri
Abstract:
One of the principal parameters defining the clay soil dynamic response is the strain-shear modulus relation. Predicting the strain and, subsequently, shear modulus reduction of the soil is essential for performance analysis of structures exposed to earthquake and dynamic loadings. Many soil properties affect soil’s dynamic behavior. In order to capture those effects, in this study, a database containing 1193 data points consists of maximum shear modulus, strain, moisture content, initial void ratio, plastic limit, liquid limit, initial confining pressure resulting from dynamic laboratory testing of 21 clays is collected for predicting the shear modulus vs. strain curve of soil. A model based on an extreme gradient boosting technique is proposed. A tree-structured parzan estimator hyper-parameter tuning algorithm is utilized simultaneously to find the best hyper-parameters for the model. The performance of the model is compared to the existing empirical equations using the coefficient of correlation and root mean square error.Keywords: XGBoost, hyper-parameter tuning, soil shear modulus, dynamic response
Procedia PDF Downloads 1953209 Predicting Oil Spills in Real-Time: A Machine Learning and AIS Data-Driven Approach
Authors: Tanmay Bisen, Aastha Shayla, Susham Biswas
Abstract:
Oil spills from tankers can cause significant harm to the environment and local communities, as well as have economic consequences. Early predictions of oil spills can help to minimize these impacts. Our proposed system uses machine learning and neural networks to predict potential oil spills by monitoring data from ship Automatic Identification Systems (AIS). The model analyzes ship movements, speeds, and changes in direction to identify patterns that deviate from the norm and could indicate a potential spill. Our approach not only identifies anomalies but also predicts spills before they occur, providing early detection and mitigation measures. This can prevent or minimize damage to the reputation of the company responsible and the country where the spill takes place. The model's performance on the MV Wakashio oil spill provides insight into its ability to detect and respond to real-world oil spills, highlighting areas for improvement and further research.Keywords: Anomaly Detection, Oil Spill Prediction, Machine Learning, Image Processing, Graph Neural Network (GNN)
Procedia PDF Downloads 643208 Resilient Machine Learning in the Nuclear Industry: Crack Detection as a Case Study
Authors: Anita Khadka, Gregory Epiphaniou, Carsten Maple
Abstract:
There is a dramatic surge in the adoption of machine learning (ML) techniques in many areas, including the nuclear industry (such as fault diagnosis and fuel management in nuclear power plants), autonomous systems (including self-driving vehicles), space systems (space debris recovery, for example), medical surgery, network intrusion detection, malware detection, to name a few. With the application of learning methods in such diverse domains, artificial intelligence (AI) has become a part of everyday modern human life. To date, the predominant focus has been on developing underpinning ML algorithms that can improve accuracy, while factors such as resiliency and robustness of algorithms have been largely overlooked. If an adversarial attack is able to compromise the learning method or data, the consequences can be fatal, especially but not exclusively in safety-critical applications. In this paper, we present an in-depth analysis of five adversarial attacks and three defence methods on a crack detection ML model. Our analysis shows that it can be dangerous to adopt machine learning techniques in security-critical areas such as the nuclear industry without rigorous testing since they may be vulnerable to adversarial attacks. While common defence methods can effectively defend against different attacks, none of the three considered can provide protection against all five adversarial attacks analysed.Keywords: adversarial machine learning, attacks, defences, nuclear industry, crack detection
Procedia PDF Downloads 1523207 Deriving Generic Transformation Matrices for Multi-Axis Milling Machine
Authors: Alan C. Lin, Tzu-Kuan Lin, Tsong Der Lin
Abstract:
This paper proposes a new method to find the equations of transformation matrix for the rotation angles of the two rotational axes and the coordinates of the three linear axes of an orthogonal multi-axis milling machine. This approach provides intuitive physical meanings for rotation angles of multi-axis machines, which can be used to evaluate the accuracy of the conversion from CL data to NC data.Keywords: CAM, multi-axis milling machining, transformation matrix, rotation angles
Procedia PDF Downloads 4723206 Possibility Theory Based Multi-Attribute Decision-Making: Application in Facility Location-Selection Problem under Uncertain and Extreme Environment
Authors: Bezhan Ghvaberidze
Abstract:
A fuzzy multi-objective facility location-selection problem (FLSP) under uncertain and extreme environments based on possibility theory is developed. The model’s uncertain parameters in the q-rung orthopair fuzzy values are presented and transformed in the Dempster-Shaper’s belief structure environment. An objective function – distribution centers’ selection ranking index as an extension of Dempster’s extremal expectations under discrimination q-rung orthopair fuzzy information is constructed. Experts evaluate each humanitarian aid from distribution centers (HADC) against each of the uncertain factors. HADCs location problem is reduced to the bicriteria problem of partitioning the set of customers by the set of centers: (1) – Minimization of transportation costs; (2) – Maximization of centers’ selection ranking indexes. Partitioning type constraints are also constructed. For an illustration of the obtained results, a numerical example is created from the facility location-selection problem.Keywords: FLSP, multi-objective combinatorial optimization problem, evidence theory, HADC, q-rung orthopair fuzzy set, possibility theory
Procedia PDF Downloads 1113205 Detecting Music Enjoyment Level Using Electroencephalogram Signals and Machine Learning Techniques
Authors: Raymond Feng, Shadi Ghiasi
Abstract:
An electroencephalogram (EEG) is a non-invasive technique that records electrical activity in the brain using scalp electrodes. Researchers have studied the use of EEG to detect emotions and moods by collecting signals from participants and analyzing how those signals correlate with their activities. In this study, researchers investigated the relationship between EEG signals and music enjoyment. Participants listened to music while data was collected. During the signal-processing phase, power spectral densities (PSDs) were computed from the signals, and dominant brainwave frequencies were extracted from the PSDs to form a comprehensive feature matrix. A machine learning approach was then taken to find correlations between the processed data and the music enjoyment level indicated by the participants. To improve on previous research, multiple machine learning models were employed, including K-Nearest Neighbors Classifier, Support Vector Classifier, and Decision Tree Classifier. Hyperparameters were used to fine-tune each model to further increase its performance. The experiments showed that a strong correlation exists, with the Decision Tree Classifier with hyperparameters yielding 85% accuracy. This study proves that EEG is a reliable means to detect music enjoyment and has future applications, including personalized music recommendation, mood adjustment, and mental health therapy.Keywords: EEG, electroencephalogram, machine learning, mood, music enjoyment, physiological signals
Procedia PDF Downloads 513204 Detecting and Thwarting Interest Flooding Attack in Information Centric Network
Authors: Vimala Rani P, Narasimha Malikarjunan, Mercy Shalinie S
Abstract:
Data Networking was brought forth as an instantiation of information-centric networking. The attackers can send a colossal number of spoofs to take hold of the Pending Interest Table (PIT) named an Interest Flooding attack (IFA) since the in- interests are recorded in the PITs of the intermediate routers until they receive corresponding Data Packets are go beyond the time limit. These attacks can be detrimental to network performance. PIT expiration rate or the Interest satisfaction rate, which cannot differentiate the IFA from attacks, is the criterion Traditional IFA detection techniques are concerned with. Threshold values can casually affect Threshold-based traditional methods. This article proposes an accurate IFA detection mechanism based on a Multiple Feature-based Extreme Learning Machine (MF-ELM). Accuracy of the attack detection can be increased by presenting the entropy of Internet names, Interest satisfaction rate and PIT usage as features extracted in the MF-ELM classifier. Furthermore, we deploy a queue-based hostile Interest prefix mitigation mechanism. The inference of this real-time test bed is that the mechanism can help the network to resist IFA with higher accuracy and efficiency.Keywords: information-centric network, pending interest table, interest flooding attack, MF-ELM classifier, queue-based mitigation strategy
Procedia PDF Downloads 1993203 Development of a Decision-Making Method by Using Machine Learning Algorithms in the Early Stage of School Building Design
Authors: Rajaian Hoonejani Mohammad, Eshraghi Pegah, Zomorodian Zahra Sadat, Tahsildoost Mohammad
Abstract:
Over the past decade, energy consumption in educational buildings has steadily increased. The purpose of this research is to provide a method to quickly predict the energy consumption of buildings using separate evaluation of zones and decomposing the building to eliminate the complexity of geometry at the early design stage. To produce this framework, machine learning algorithms such as Support vector regression (SVR) and Artificial neural network (ANN) are used to predict energy consumption and thermal comfort metrics in a school as a case. The database consists of more than 55000 samples in three climates of Iran. Cross-validation evaluation and unseen data have been used for validation. In a specific label, cooling energy, it can be said the accuracy of prediction is at least 84% and 89% in SVR and ANN, respectively. The results show that the SVR performed much better than the ANN.Keywords: early stage of design, energy, thermal comfort, validation, machine learning
Procedia PDF Downloads 623202 Autonomous Kuka Youbot Navigation Based on Machine Learning and Path Planning
Authors: Carlos Gordon, Patricio Encalada, Henry Lema, Diego Leon, Dennis Chicaiza
Abstract:
The following work presents a proposal of autonomous navigation of mobile robots implemented in an omnidirectional robot Kuka Youbot. We have been able to perform the integration of robotic operative system (ROS) and machine learning algorithms. ROS mainly provides two distributions; ROS hydro and ROS Kinect. ROS hydro allows managing the nodes of odometry, kinematics, and path planning with statistical and probabilistic, global and local algorithms based on Adaptive Monte Carlo Localization (AMCL) and Dijkstra. Meanwhile, ROS Kinect is responsible for the detection block of dynamic objects which can be in the points of the planned trajectory obstructing the path of Kuka Youbot. The detection is managed by artificial vision module under a trained neural network based on the single shot multibox detector system (SSD), where the main dynamic objects for detection are human beings and domestic animals among other objects. When the objects are detected, the system modifies the trajectory or wait for the decision of the dynamic obstacle. Finally, the obstacles are skipped from the planned trajectory, and the Kuka Youbot can reach its goal thanks to the machine learning algorithms.Keywords: autonomous navigation, machine learning, path planning, robotic operative system, open source computer vision library
Procedia PDF Downloads 1713201 Application of Supervised Deep Learning-based Machine Learning to Manage Smart Homes
Authors: Ahmed Al-Adaileh
Abstract:
Renewable energy sources, domestic storage systems, controllable loads and machine learning technologies will be key components of future smart homes management systems. An energy management scheme that uses a Deep Learning (DL) approach to support the smart home management systems, which consist of a standalone photovoltaic system, storage unit, heating ventilation air-conditioning system and a set of conventional and smart appliances, is presented. The objective of the proposed scheme is to apply DL-based machine learning to predict various running parameters within a smart home's environment to achieve maximum comfort levels for occupants, reduced electricity bills, and less dependency on the public grid. The problem is using Reinforcement learning, where decisions are taken based on applying the Continuous-time Markov Decision Process. The main contribution of this research is the proposed framework that applies DL to enhance the system's supervised dataset to offer unlimited chances to effectively support smart home systems. A case study involving a set of conventional and smart appliances with dedicated processing units in an inhabited building can demonstrate the validity of the proposed framework. A visualization graph can show "before" and "after" results.Keywords: smart homes systems, machine learning, deep learning, Markov Decision Process
Procedia PDF Downloads 1873200 Predictive Modeling of Student Behavior in Virtual Reality: A Machine Learning Approach
Authors: Gayathri Sadanala, Shibam Pokhrel, Owen Murphy
Abstract:
In the ever-evolving landscape of education, Virtual Reality (VR) environments offer a promising avenue for enhancing student engagement and learning experiences. However, understanding and predicting student behavior within these immersive settings remain challenging tasks. This paper presents a comprehensive study on the predictive modeling of student behavior in VR using machine learning techniques. We introduce a rich data set capturing student interactions, movements, and progress within a VR orientation program. The dataset is divided into training and testing sets, allowing us to develop and evaluate predictive models for various aspects of student behavior, including engagement levels, task completion, and performance. Our machine learning approach leverages a combination of feature engineering and model selection to reveal hidden patterns in the data. We employ regression and classification models to predict student outcomes, and the results showcase promising accuracy in forecasting behavior within VR environments. Furthermore, we demonstrate the practical implications of our predictive models for personalized VR-based learning experiences and early intervention strategies. By uncovering the intricate relationship between student behavior and VR interactions, we provide valuable insights for educators, designers, and developers seeking to optimize virtual learning environments.Keywords: interaction, machine learning, predictive modeling, virtual reality
Procedia PDF Downloads 1233199 Automated Detection of Women Dehumanization in English Text
Authors: Maha Wiss, Wael Khreich
Abstract:
Animals, objects, foods, plants, and other non-human terms are commonly used as a source of metaphors to describe females in formal and slang language. Comparing women to non-human items not only reflects cultural views that might conceptualize women as subordinates or in a lower position than humans, yet it conveys this degradation to the listeners. Moreover, the dehumanizing representation of females in the language normalizes the derogation and even encourages sexism and aggressiveness against women. Although dehumanization has been a popular research topic for decades, according to our knowledge, no studies have linked women's dehumanizing language to the machine learning field. Therefore, we introduce our research work as one of the first attempts to create a tool for the automated detection of the dehumanizing depiction of females in English texts. We also present the first labeled dataset on the charted topic, which is used for training supervised machine learning algorithms to build an accurate classification model. The importance of this work is that it accomplishes the first step toward mitigating dehumanizing language against females.Keywords: gender bias, machine learning, NLP, women dehumanization
Procedia PDF Downloads 743198 Hand Gesture Interpretation Using Sensing Glove Integrated with Machine Learning Algorithms
Authors: Aqsa Ali, Aleem Mushtaq, Attaullah Memon, Monna
Abstract:
In this paper, we present a low cost design for a smart glove that can perform sign language recognition to assist the speech impaired people. Specifically, we have designed and developed an Assistive Hand Gesture Interpreter that recognizes hand movements relevant to the American Sign Language (ASL) and translates them into text for display on a Thin-Film-Transistor Liquid Crystal Display (TFT LCD) screen as well as synthetic speech. Linear Bayes Classifiers and Multilayer Neural Networks have been used to classify 11 feature vectors obtained from the sensors on the glove into one of the 27 ASL alphabets and a predefined gesture for space. Three types of features are used; bending using six bend sensors, orientation in three dimensions using accelerometers and contacts at vital points using contact sensors. To gauge the performance of the presented design, the training database was prepared using five volunteers. The accuracy of the current version on the prepared dataset was found to be up to 99.3% for target user. The solution combines electronics, e-textile technology, sensor technology, embedded system and machine learning techniques to build a low cost wearable glove that is scrupulous, elegant and portable.Keywords: American sign language, assistive hand gesture interpreter, human-machine interface, machine learning, sensing glove
Procedia PDF Downloads 2913197 Analyzing the Performance of Machine Learning Models to Predict Alzheimer's Disease and its Stages Addressing Missing Value Problem
Authors: Carlos Theran, Yohn Parra Bautista, Victor Adankai, Richard Alo, Jimwi Liu, Clement G. Yedjou
Abstract:
Alzheimer's disease (AD) is a neurodegenerative disorder primarily characterized by deteriorating cognitive functions. AD has gained relevant attention in the last decade. An estimated 24 million people worldwide suffered from this disease by 2011. In 2016 an estimated 40 million were diagnosed with AD, and for 2050 is expected to reach 131 million people affected by AD. Therefore, detecting and confirming AD at its different stages is a priority for medical practices to provide adequate and accurate treatments. Recently, Machine Learning (ML) models have been used to study AD's stages handling missing values in multiclass, focusing on the delineation of Early Mild Cognitive Impairment (EMCI), Late Mild Cognitive Impairment (LMCI), and normal cognitive (CN). But, to our best knowledge, robust performance information of these models and the missing data analysis has not been presented in the literature. In this paper, we propose studying the performance of five different machine learning models for AD's stages multiclass prediction in terms of accuracy, precision, and F1-score. Also, the analysis of three imputation methods to handle the missing value problem is presented. A framework that integrates ML model for AD's stages multiclass prediction is proposed, performing an average accuracy of 84%.Keywords: alzheimer's disease, missing value, machine learning, performance evaluation
Procedia PDF Downloads 2383196 Fraud Detection in Credit Cards with Machine Learning
Authors: Anjali Chouksey, Riya Nimje, Jahanvi Saraf
Abstract:
Online transactions have increased dramatically in this new ‘social-distancing’ era. With online transactions, Fraud in online payments has also increased significantly. Frauds are a significant problem in various industries like insurance companies, baking, etc. These frauds include leaking sensitive information related to the credit card, which can be easily misused. Due to the government also pushing online transactions, E-commerce is on a boom. But due to increasing frauds in online payments, these E-commerce industries are suffering a great loss of trust from their customers. These companies are finding credit card fraud to be a big problem. People have started using online payment options and thus are becoming easy targets of credit card fraud. In this research paper, we will be discussing machine learning algorithms. We have used a decision tree, XGBOOST, k-nearest neighbour, logistic-regression, random forest, and SVM on a dataset in which there are transactions done online mode using credit cards. We will test all these algorithms for detecting fraud cases using the confusion matrix, F1 score, and calculating the accuracy score for each model to identify which algorithm can be used in detecting frauds.Keywords: machine learning, fraud detection, artificial intelligence, decision tree, k nearest neighbour, random forest, XGBOOST, logistic regression, support vector machine
Procedia PDF Downloads 1383195 Automatic Aggregation and Embedding of Microservices for Optimized Deployments
Authors: Pablo Chico De Guzman, Cesar Sanchez
Abstract:
Microservices are a software development methodology in which applications are built by composing a set of independently deploy-able, small, modular services. Each service runs a unique process and it gets instantiated and deployed in one or more machines (we assume that different microservices are deployed into different machines). Microservices are becoming the de facto standard for developing distributed cloud applications due to their reduced release cycles. In principle, the responsibility of a microservice can be as simple as implementing a single function, which can lead to the following issues: - Resource fragmentation due to the virtual machine boundary. - Poor communication performance between microservices. Two composition techniques can be used to optimize resource fragmentation and communication performance: aggregation and embedding of microservices. Aggregation allows the deployment of a set of microservices on the same machine using a proxy server. Aggregation helps to reduce resource fragmentation, and is particularly useful when the aggregated services have a similar scalability behavior. Embedding deals with communication performance by deploying on the same virtual machine those microservices that require a communication channel (localhost bandwidth is reported to be about 40 times faster than cloud vendor local networks and it offers better reliability). Embedding can also reduce dependencies on load balancer services since the communication takes place on a single virtual machine. For example, assume that microservice A has two instances, a1 and a2, and it communicates with microservice B, which also has two instances, b1 and b2. One embedding can deploy a1 and b1 on machine m1, and a2 and b2 are deployed on a different machine m2. This deployment configuration allows each pair (a1-b1), (a2-b2) to communicate using the localhost interface without the need of a load balancer between microservices A and B. Aggregation and embedding techniques are complex since different microservices might have incompatible runtime dependencies which forbid them from being installed on the same machine. There is also a security concern since the attack surface between microservices can be larger. Luckily, container technology allows to run several processes on the same machine in an isolated manner, solving the incompatibility of running dependencies and the previous security concern, thus greatly simplifying aggregation/embedding implementations by just deploying a microservice container on the same machine as the aggregated/embedded microservice container. Therefore, a wide variety of deployment configurations can be described by combining aggregation and embedding to create an efficient and robust microservice architecture. This paper presents a formal method that receives a declarative definition of a microservice architecture and proposes different optimized deployment configurations by aggregating/embedding microservices. The first prototype is based on i2kit, a deployment tool also submitted to ICWS 2018. The proposed prototype optimizes the following parameters: network/system performance, resource usage, resource costs and failure tolerance.Keywords: aggregation, deployment, embedding, resource allocation
Procedia PDF Downloads 1973194 Grating Scale Thermal Expansion Error Compensation for Large Machine Tools Based on Multiple Temperature Detection
Authors: Wenlong Feng, Zhenchun Du, Jianguo Yang
Abstract:
To decrease the grating scale thermal expansion error, a novel method which based on multiple temperature detections is proposed. Several temperature sensors are installed on the grating scale and the temperatures of these sensors are recorded. The temperatures of every point on the grating scale are calculated by interpolating between adjacent sensors. According to the thermal expansion principle, the grating scale thermal expansion error model can be established by doing the integral for the variations of position and temperature. A novel compensation method is proposed in this paper. By applying the established error model, the grating scale thermal expansion error is decreased by 90% compared with no compensation. The residual positioning error of the grating scale is less than 15um/10m and the accuracy of the machine tool is significant improved.Keywords: thermal expansion error of grating scale, error compensation, machine tools, integral method
Procedia PDF Downloads 3603193 Regression Model Evaluation on Depth Camera Data for Gaze Estimation
Authors: James Purnama, Riri Fitri Sari
Abstract:
We investigate the machine learning algorithm selection problem in the term of a depth image based eye gaze estimation, with respect to its essential difficulty in reducing the number of required training samples and duration time of training. Statistics based prediction accuracy are increasingly used to assess and evaluate prediction or estimation in gaze estimation. This article evaluates Root Mean Squared Error (RMSE) and R-Squared statistical analysis to assess machine learning methods on depth camera data for gaze estimation. There are 4 machines learning methods have been evaluated: Random Forest Regression, Regression Tree, Support Vector Machine (SVM), and Linear Regression. The experiment results show that the Random Forest Regression has the lowest RMSE and the highest R-Squared, which means that it is the best among other methods.Keywords: gaze estimation, gaze tracking, eye tracking, kinect, regression model, orange python
Procedia PDF Downloads 5313192 Innovative Predictive Modeling and Characterization of Composite Material Properties Using Machine Learning and Genetic Algorithms
Authors: Hamdi Beji, Toufik Kanit, Tanguy Messager
Abstract:
This study aims to construct a predictive model proficient in foreseeing the linear elastic and thermal characteristics of composite materials, drawing on a multitude of influencing parameters. These parameters encompass the shape of inclusions (circular, elliptical, square, triangle), their spatial coordinates within the matrix, orientation, volume fraction (ranging from 0.05 to 0.4), and variations in contrast (spanning from 10 to 200). A variety of machine learning techniques are deployed, including decision trees, random forests, support vector machines, k-nearest neighbors, and an artificial neural network (ANN), to facilitate this predictive model. Moreover, this research goes beyond the predictive aspect by delving into an inverse analysis using genetic algorithms. The intent is to unveil the intrinsic characteristics of composite materials by evaluating their thermomechanical responses. The foundation of this research lies in the establishment of a comprehensive database that accounts for the array of input parameters mentioned earlier. This database, enriched with this diversity of input variables, serves as a bedrock for the creation of machine learning and genetic algorithm-based models. These models are meticulously trained to not only predict but also elucidate the mechanical and thermal conduct of composite materials. Remarkably, the coupling of machine learning and genetic algorithms has proven highly effective, yielding predictions with remarkable accuracy, boasting scores ranging between 0.97 and 0.99. This achievement marks a significant breakthrough, demonstrating the potential of this innovative approach in the field of materials engineering.Keywords: machine learning, composite materials, genetic algorithms, mechanical and thermal proprieties
Procedia PDF Downloads 523191 Regionalization of IDF Curves with L-Moments for Storm Events
Authors: Noratiqah Mohd Ariff, Abdul Aziz Jemain, Mohd Aftar Abu Bakar
Abstract:
The construction of Intensity-Duration-Frequency (IDF) curves is one of the most common and useful tools in order to design hydraulic structures and to provide a mathematical relationship between rainfall characteristics. IDF curves, especially those in Peninsular Malaysia, are often built using moving windows of rainfalls. However, these windows do not represent the actual rainfall events since the duration of rainfalls is usually prefixed. Hence, instead of using moving windows, this study aims to find regionalized distributions for IDF curves of extreme rainfalls based on storm events. Homogeneity test is performed on annual maximum of storm intensities to identify homogeneous regions of storms in Peninsular Malaysia. The L-moment method is then used to regionalized Generalized Extreme Value (GEV) distribution of these annual maximums and subsequently. IDF curves are constructed using the regional distributions. The differences between the IDF curves obtained and IDF curves found using at-site GEV distributions are observed through the computation of the coefficient of variation of root mean square error, mean percentage difference and the coefficient of determination. The small differences implied that the construction of IDF curves could be simplified by finding a general probability distribution of each region. This will also help in constructing IDF curves for sites with no rainfall station.Keywords: IDF curves, L-moments, regionalization, storm events
Procedia PDF Downloads 5193190 The Relationship between Organization Culture and Organization Learning in Three Different Types of Companies
Authors: Mahmoud Timar, Javad Joukar Borazjani
Abstract:
A dynamic organization helps the management to overcome both internal and external uncertainties and complexities of the organization with more confidence and efficiency. Regarding this issue, in this paper, the influence of organizational culture factors over organizational learning components, which both of them are considered as important characteristics of a dynamic organization, has been studied in three subsidiary companies (production, consultation and service) of National Iranian Oil Company, and moreover we also tried to identify the most dominant culture in these three subsidiaries. Analysis of 840 received questionnaires by SPSS shows that there is a significant relationship between the components of organizational culture and organizational learning; however the rate of relationship between these two factors was different among the examined companies. By the use of Regression, it has been clarified that in the servicing company the highest relationship is between mission and learning environment, while in production division, there is a significant relationship between adaptability and learning needs satisfaction and however in consulting company the highest relationship is between involvement and applying learning in workplace.Keywords: denison model, culture, leaning, organizational culture, organizational learning
Procedia PDF Downloads 3663189 Value at Risk and Expected Shortfall of Firms in the Main European Union Stock Market Indexes: A Detailed Analysis by Economic Sectors and Geographical Situation
Authors: Emma M. Iglesias
Abstract:
We have analyzed extreme movements of the main stocks traded in the Eurozone in the 2000-2012 period. Our results can help future very-risk-averse investors to choose their portfolios in the Eurozone for risk management purposes. We find two main results. First, we can clearly classify firms by economic sector according to their different estimated VaR values in five of the seven countries we analyze. In special, we find sectors in general where companies have very high (telecommunications and banking) and very low (petroleum, utilities, energy and consumption) estimated VaR values. Second, we only find differences according to the geographical situation of where the stocks are traded in two countries: (1) all firms in the Irish stock market (the only financially rescued country we analyze) have very high estimated VaR values in all sectors; while (2) in Spain all firms have very low estimated VaR values including in the banking and the telecommunications sectors. All our results are supported when we study also the expected shortfall of the firms.Keywords: risk management, firms, pareto tail thickness parameter, GARCH-type models, value-at-risk, extreme value theory, heavy tails, stock indexes, eurozone
Procedia PDF Downloads 3643188 Comparing Deep Architectures for Selecting Optimal Machine Translation
Authors: Despoina Mouratidis, Katia Lida Kermanidis
Abstract:
Machine translation (MT) is a very important task in Natural Language Processing (NLP). MT evaluation is crucial in MT development, as it constitutes the means to assess the success of an MT system, and also helps improve its performance. Several methods have been proposed for the evaluation of (MT) systems. Some of the most popular ones in automatic MT evaluation are score-based, such as the BLEU score, and others are based on lexical similarity or syntactic similarity between the MT outputs and the reference involving higher-level information like part of speech tagging (POS). This paper presents a language-independent machine learning framework for classifying pairwise translations. This framework uses vector representations of two machine-produced translations, one from a statistical machine translation model (SMT) and one from a neural machine translation model (NMT). The vector representations consist of automatically extracted word embeddings and string-like language-independent features. These vector representations used as an input to a multi-layer neural network (NN) that models the similarity between each MT output and the reference, as well as between the two MT outputs. To evaluate the proposed approach, a professional translation and a "ground-truth" annotation are used. The parallel corpora used are English-Greek (EN-GR) and English-Italian (EN-IT), in the educational domain and of informal genres (video lecture subtitles, course forum text, etc.) that are difficult to be reliably translated. They have tested three basic deep learning (DL) architectures to this schema: (i) fully-connected dense, (ii) Convolutional Neural Network (CNN), and (iii) Long Short-Term Memory (LSTM). Experiments show that all tested architectures achieved better results when compared against those of some of the well-known basic approaches, such as Random Forest (RF) and Support Vector Machine (SVM). Better accuracy results are obtained when LSTM layers are used in our schema. In terms of a balance between the results, better accuracy results are obtained when dense layers are used. The reason for this is that the model correctly classifies more sentences of the minority class (SMT). For a more integrated analysis of the accuracy results, a qualitative linguistic analysis is carried out. In this context, problems have been identified about some figures of speech, as the metaphors, or about certain linguistic phenomena, such as per etymology: paronyms. It is quite interesting to find out why all the classifiers led to worse accuracy results in Italian as compared to Greek, taking into account that the linguistic features employed are language independent.Keywords: machine learning, machine translation evaluation, neural network architecture, pairwise classification
Procedia PDF Downloads 1233187 Climate Change Threats to UNESCO-Designated World Heritage Sites: Empirical Evidence from Konso Cultural Landscape, Ethiopia
Authors: Yimer Mohammed Assen, Abiyot Legesse Kura, Engida Esyas Dube, Asebe Regassa Debelo, Girma Kelboro Mensuro, Lete Bekele Gure
Abstract:
Climate change has posed severe threats to many cultural landscapes of UNESCO world heritage sites recently. The UNESCO State of Conservation (SOC) reports categorized flooding, temperature increment, and drought as threats to cultural landscapes. This study aimed to examine variations and trends of rainfall and temperature extreme events and their threats to the UNESCO-designated Konso Cultural Landscape in southern Ethiopia. The study used dense merged satellite-gauge station rainfall data (1981-2020) with spatial resolution of 4km by 4km and observed maximum and minimum temperature data (1987-2020). Qualitative data were also gathered from cultural leaders, local administrators, and religious leaders using structured interview checklists. The spatial patterns, coefficient of variation, standardized anomalies, trends, and magnitude of change of rainfall and temperature extreme events both at annual and seasonal levels were computed using the Mann-Kendall trend test and Sen’s slope estimator under the CDT package. The standard precipitation index (SPI) was also used to calculate drought severity, frequency, and trend maps. The data gathered from key informant interviews and focus group discussions were coded and analyzed thematically to complement statistical findings. Thematic areas that explain the impacts of extreme events on the cultural landscape were chosen for coding. The thematic analysis was conducted using Nvivo software. The findings revealed that rainfall was highly variable and unpredictable, resulting in extreme drought and flood. There were significant (P<0.05) increasing trends of heavy rainfall (R10mm and R20mm) and the total amount of rain on wet days (PRCPTOT), which might have resulted in flooding. The study also confirmed that absolute temperature extreme indices (TXx, TXn, and TNx) and the percentile-based temperature extreme indices (TX90p, TN90p, TX10p, and TN10P) showed significant (P<0.05) increasing trends which are signals for warming of the study area. The results revealed that the frequency as well as the severity of drought at 3-months (katana/hageya seasons) was more pronounced than the 12-months (annual) time scale. The highest number of droughts in 100 years is projected at a 3-months timescale across the study area. The findings also showed that frequent drought has led to loss of grasses which are used for making traditional individual houses and multipurpose communal houses (pafta), food insecurity, migration, loss of biodiversity, and commodification of stones from terrace. On the other hand, the increasing trends of rainfall extreme indices resulted in destruction of terraces, soil erosion, loss of life and damage of properties. The study shows that a persistent decline in farmland productivity, due to erratic and extreme rainfall and frequent drought occurrences, forced the local people to participate in non-farm activities and retreat from daily preservation and management of their landscape. Overall, the increasing rainfall and temperature extremes coupled with prevalence of drought are thought to have an impact on the sustainability of cultural landscape through disrupting the ecosystem services and livelihood of the community. Therefore, more localized adaptation and mitigation strategies to the changing climate are needed to maintain the sustainability of Konso cultural landscapes as a global cultural treasure and to strengthen the resilience of smallholder farmers.Keywords: adaptation, cultural landscape, drought, extremes indices
Procedia PDF Downloads 173186 Assessment of Ultra-High Cycle Fatigue Behavior of EN-GJL-250 Cast Iron Using Ultrasonic Fatigue Testing Machine
Authors: Saeedeh Bakhtiari, Johannes Depessemier, Stijn Hertelé, Wim De Waele
Abstract:
High cycle fatigue comprising up to 107 load cycles has been the subject of many studies, and the behavior of many materials was recorded adequately in this regime. However, many applications involve larger numbers of load cycles during the lifetime of machine components. In this ultra-high cycle regime, other failure mechanisms play, and the concept of a fatigue endurance limit (assumed for materials such as steel) is often an oversimplification of reality. When machine component design demands a high geometrical complexity, cast iron grades become interesting candidate materials. Grey cast iron is known for its low cost, high compressive strength, and good damping properties. However, the ultra-high cycle fatigue behavior of cast iron is poorly documented. The current work focuses on the ultra-high cycle fatigue behavior of EN-GJL-250 (GG25) grey cast iron by developing an ultrasonic (20 kHz) fatigue testing system. Moreover, the testing machine is instrumented to measure the temperature and the displacement of the specimen, and to control the temperature. The high resonance frequency allowed to assess the behavior of the cast iron of interest within a matter of days for ultra-high numbers of cycles, and repeat the tests to quantify the natural scatter in fatigue resistance.Keywords: GG25, cast iron, ultra-high cycle fatigue, ultrasonic test
Procedia PDF Downloads 1683185 Machine Learning for Targeting of Conditional Cash Transfers: Improving the Effectiveness of Proxy Means Tests to Identify Future School Dropouts and the Poor
Authors: Cristian Crespo
Abstract:
Conditional cash transfers (CCTs) have been targeted towards the poor. Thus, their targeting assessments check whether these schemes have been allocated to low-income households or individuals. However, CCTs have more than one goal and target group. An additional goal of CCTs is to increase school enrolment. Hence, students at risk of dropping out of school also are a target group. This paper analyses whether one of the most common targeting mechanisms of CCTs, a proxy means test (PMT), is suitable to identify the poor and future school dropouts. The PMT is compared with alternative approaches that use the outputs of a predictive model of school dropout. This model was built using machine learning algorithms and rich administrative datasets from Chile. The paper shows that using machine learning outputs in conjunction with the PMT increases targeting effectiveness by identifying more students who are either poor or future dropouts. This joint targeting approach increases effectiveness in different scenarios except when the social valuation of the two target groups largely differs. In these cases, the most likely optimal approach is to solely adopt the targeting mechanism designed to find the highly valued group.Keywords: conditional cash transfers, machine learning, poverty, proxy means tests, school dropout prediction, targeting
Procedia PDF Downloads 1953184 Computational Intelligence and Machine Learning for Urban Drainage Infrastructure Asset Management
Authors: Thewodros K. Geberemariam
Abstract:
The rapid physical expansion of urbanization coupled with aging infrastructure presents a unique decision and management challenges for many big city municipalities. Cities must therefore upgrade and maintain the existing aging urban drainage infrastructure systems to keep up with the demands. Given the overall contribution of assets to municipal revenue and the importance of infrastructure to the success of a livable city, many municipalities are currently looking for a robust and smart urban drainage infrastructure asset management solution that combines management, financial, engineering and technical practices. This robust decision-making shall rely on sound, complete, current and relevant data that enables asset valuation, impairment testing, lifecycle modeling, and forecasting across the multiple asset portfolios. On this paper, predictive computational intelligence (CI) and multi-class machine learning (ML) coupled with online, offline, and historical record data that are collected from an array of multi-parameter sensors are used for the extraction of different operational and non-conforming patterns hidden in structured and unstructured data to determine and produce actionable insight on the current and future states of the network. This paper aims to improve the strategic decision-making process by identifying all possible alternatives; evaluate the risk of each alternative, and choose the alternative most likely to attain the required goal in a cost-effective manner using historical and near real-time urban drainage infrastructure data for urban drainage infrastructures assets that have previously not benefited from computational intelligence and machine learning advancements.Keywords: computational intelligence, machine learning, urban drainage infrastructure, machine learning, classification, prediction, asset management space
Procedia PDF Downloads 1463183 Design of Fuzzy Logic Based Global Power System Stabilizer for Dynamic Stability Enhancement in Multi-Machine Power System
Authors: N. P. Patidar, J. Earnest, Laxmikant Nagar, Akshay Sharma
Abstract:
This paper describes the diligence of a new input signal based fuzzy power system stabilizer in multi-machine power system. Instead of conventional input pairs like speed deviation (∆ω) and derivative of speed deviation i.e. acceleration (∆ω ̇) or speed deviation and accelerating power deviation of each machine, in this paper, deviation of active power through the tie line colligating two areas is used as one of the inputs to the fuzzy logic controller in concurrence with the speed deviation. Fuzzy Logic has the features of simple concept, easy effectuation, and computationally efficient. The advantage of this input is that, the same signal can be fed to each of the fuzzy logic controller connected with each machine. The simulated system comprises of two fully symmetrical areas coupled together by two 230 kV lines. Each area is equipped with two superposable generators rated 20 kV/900MVA and area-1 is exporting 413 MW to area-2. The effectiveness of the proposed control scheme has been assessed by performing small signal stability assessment and transient stability assessment. The proposed control scheme has been compared with a conventional PSS. Digital simulation is used to demonstrate the performance of fuzzy logic controller.Keywords: Power System Stabilizer (PSS), small signal stability, inter-area oscillation, fuzzy logic controller, membership function, rule base
Procedia PDF Downloads 521