Search results for: Artificial Neural Network (ANNs)

Authors: Thamer Mahmmoud Muhammad Al Jbarat

Abstract:

Building Performance Simulation plays a crucial role in optimizing energy efficiency, comfort, and sustainability in buildings. This paper explores the integration of Artificial Intelligence techniques into Building Performance Simulation to enhance accuracy, efficiency, and adaptability. The synthesis of Artificial Intelligence and Building Performance Simulation offers promising avenues for addressing complex building dynamics, optimizing energy consumption, and improving occupants' comfort. This paper examines various Artificial Intelligence methodologies and their applications in Building Performance Simulation, highlighting their potential benefits and challenges. Through a comprehensive review of existing literature and case studies, this paper presents insights into the current state, future directions, and implications of Artificial Intelligence driven Building Performance Simulation on the built environment

Keywords: artificial intelligence, building performance, energy efficiency, building performance simulation, buildings sustainability, built environment.

Procedia PDF Downloads 26

Authors: Boguslaw Schreyer

Abstract:

The paper investigates the application of the Kohonen’s Self-Organizing Map (SOM) to the wheeled robot starting and braking dynamic states. In securing wheeled robot stability as well as minimum starting and braking time, it is important to ensure correct torque distribution as well as proper slope of braking and driving moments. In this paper, a correct movement distribution has been formulated, securing optimum adhesion coefficient and good transversal stability of a wheeled robot. A neural tuner has been proposed to secure the above properties, although most of the attention is attached to the SOM network application. If the delay of the torque application or torque release is not negligible, it is important to change the rising and falling slopes of the torque. The road/surface condition is also paramount in robot dynamic states control. As the road conditions may randomly change in time, application of the SOM network has been suggested in order to classify the actual road conditions.

Keywords: slip control, SOM network, torque distribution, wheeled Robot

Procedia PDF Downloads 126

Authors: Parthiban Srinivasan

Abstract:

With recent advances in computational power and access to enough data in biosciences, artificial intelligence methods are increasingly being used in drug discovery research. These methods are essentially a series of advanced statistics based exercises that review the past to indicate the likely future. Our goal is to develop a model that accurately predicts biological activity and toxicity parameters for novel compounds. We have compiled a robust library of over 150,000 chemical compounds with different pharmacological properties from literature and public domain databases. The compounds are stored in simplified molecular-input line-entry system (SMILES), a commonly used text encoding for organic molecules. We utilize an automated process to generate an array of numerical descriptors (features) for each molecule. Redundant and irrelevant descriptors are eliminated iteratively. Our prediction engine is based on a portfolio of machine learning algorithms. We found Random Forest algorithm to be a better choice for this analysis. We captured non-linear relationship in the data and formed a prediction model with reasonable accuracy by averaging across a large number of randomized decision trees. Our next step is to apply deep neural network (DNN) algorithm to predict the biological activity and toxicity properties. We expect the DNN algorithm to give better results and improve the accuracy of the prediction. This presentation will review all these prominent machine learning and deep learning methods, our implementation protocols and discuss these techniques for their usefulness in biomedical and health informatics.

Keywords: deep learning, drug discovery, health informatics, machine learning, toxicity prediction

Procedia PDF Downloads 357

Authors: Deokho Kim, Wonwoo Lee, Jaewoong Lee, Teresa Ng, Gun-Ill Lee, Jiwon Jeong

Abstract:

This paper presents a method for improving object search accuracy using a deep learning model. A major limitation to provide accurate similarity with deep learning is the requirement of huge amount of data for training pairwise similarity scores (metrics), which is impractical to collect. Thus, similarity scores are usually trained with a relatively small dataset, which comes from a different domain, causing limited accuracy on measuring similarity. For this reason, this paper proposes a deep learning model that can be trained with a significantly small amount of data, a clustered data which of each cluster contains a set of visually similar images. In order to measure similarity distance with the proposed method, visual features of two images are extracted from intermediate layers of a convolutional neural network with various pooling methods, and the network is trained with pairwise similarity scores which is defined zero for images in identical cluster. The proposed method outperforms the state-of-the-art object similarity scoring techniques on evaluation for finding exact items. The proposed method achieves 86.5% of accuracy compared to the accuracy of the state-of-the-art technique, which is 59.9%. That is, an exact item can be found among four retrieved images with an accuracy of 86.5%, and the rest can possibly be similar products more than the accuracy. Therefore, the proposed method can greatly reduce the amount of training data with an order of magnitude as well as providing a reliable similarity metric.

Keywords: visual search, deep learning, convolutional neural network, machine learning

Procedia PDF Downloads 215

Authors: Furkan Ozkan, M. Selcuk Arslan, Hatice Mercan

Abstract:

Electric vehicles are becoming increasingly popular asa sustainable alternative to traditional combustion engine vehicles. However, to fully realize the potential of EVs in reducing environmental impact and energy consumption, efficient control strategies are essential. This study explores the application of green wave control using model predictive control for electric vehicles, coupled with energy consumption modeling using neural networks. The use of MPC allows for real-time optimization of the vehicles’ energy consumption while considering dynamic traffic conditions. By leveraging neural networks for energy consumption modeling, the EV's performance can be further enhanced through accurate predictions and adaptive control. The integration of these advanced control and modeling techniques aims to maximize energy efficiency and range while navigating urban traffic scenarios. The findings of this research offer valuable insights into the potential of green wave control for electric vehicles and demonstrate the significance of integrating MPC and neural network modeling for optimizing energy consumption. This work contributes to the advancement of sustainable transportation systems and the widespread adoption of electric vehicles. To evaluate the effectiveness of the green wave control strategy in real-world urban environments, extensive simulations were conducted using a high-fidelity vehicle model and realistic traffic scenarios. The results indicate that the integration of model predictive control and energy consumption modeling with neural networks had a significant impact on the energy efficiency and range of electric vehicles. Through the use of MPC, the electric vehicle was able to adapt its speed and acceleration profile in realtime to optimize energy consumption while maintaining travel time objectives. The neural network-based energy consumption modeling provided accurate predictions, enabling the vehicle to anticipate and respond to variations in traffic flow, further enhancing energy efficiency and range. Furthermore, the study revealed that the green wave control strategy not only reduced energy consumption but also improved the overall driving experience by minimizing abrupt acceleration and deceleration, leading to a smoother and more comfortable ride for passengers. These results demonstrate the potential for green wave control to revolutionize urban transportation by enhancing the performance of electric vehicles and contributing to a more sustainable and efficient mobility ecosystem.

Keywords: electric vehicles, energy efficiency, green wave control, model predictive control, neural networks

Procedia PDF Downloads 54

Authors: Abebe Mulugojam Negash, Yongbin Yu, Ekong Favour, Bekalu Nigus Dawit, Molla Woretaw Teshome, Aynalem Birtukan Yirga

Abstract:

Breast cancer remains a significant health concern, necessitating advancements in diagnostic methodologies. Addressing this, our paper confronts the notable challenges in breast cancer classification, particularly the imbalance in datasets and the constraints in the accuracy and interpretability of prevailing deep learning approaches. We proposed an attention-based residual neural network (ResNet), which effectively combines the robust features of ResNet with an advanced attention mechanism. Enhanced through strategic data augmentation and positive weight adjustments, this approach specifically targets the issue of data imbalance. The proposed model is tested on the BreakHis dataset and achieved accuracies of 99.00%, 99.04%, 98.67%, and 98.08% in different magnifications (40X, 100X, 200X, and 400X), respectively. We evaluated the performance by using different evaluation metrics such as precision, recall, and F1-Score and made comparisons with other state-of-the-art methods. Our experiments demonstrate that the proposed model outperforms existing approaches, achieving higher accuracy in breast cancer classification.

Keywords: residual neural network, attention mechanism, positive weight, data augmentation

Procedia PDF Downloads 101

Authors: Adriana Martínez-Arias, John Josephraj Selvaraj, Luis Octavio González-Salcedo

Abstract:

In the Colombian Pacific, Dorado fish (Coryphaena hippurus L.) fisheries is of great commercial interest. However, its habitat and fisheries may be affected by climatic change especially by the actual increase in sea surface temperature. Hence, it is of interest to study the dynamics of these species fishing zones. In this study, we developed Artificial Neural Networks (ANN) models to predict Catch per Unit Effort (CPUE) as an indicator of species abundance. The model was based on four oceanographic variables (Chlorophyll a, Sea Surface Temperature, Sea Level Anomaly and Bathymetry) derived from satellite data. CPUE datasets for model training and cross-validation were obtained from logbooks of commercial fishing vessel. Sea surface Temperature for Colombian Pacific were projected under Representative Concentration Pathway (RCP) scenarios 4.5 and 8.5 using Coupled Model Intercomparison Project Phase 5 (CMIP5) and CPUE maps were created. Our results indicated that an increase in sea surface temperature reduces the potential fishing zones of this species in the Colombian Pacific. We conclude that ANN is a reliable tool for simulation of climate change effects on the potential fishing zones. This research opens a future agenda for other species that have been affected by climate change.

Keywords: climatic change, artificial neural networks, dorado fish, CPUE

Procedia PDF Downloads 243

Authors: Kaouther Selmi, Alaeddine Sridi, Mohamed Bouallegue, Kais Bouallegue

Abstract:

Neurons are the fundamental units of the brain and the nervous system. The variable structure model of neurons consists of a system of differential equations with various parameters. By optimizing these parameters, we can create a unique model that describes the dynamic behavior of a single neuron. We introduce a neural network based on neurons with multiple dendrites employing an activation function with a variable structure. In this paper, we present a model for heart behavior. Finally, we showcase our successful simulation of the heart's ECG diagram using our Variable Structure Neuron Model (VSMN). This result could provide valuable insights into cardiology.

Keywords: neural networks, neuron, dendrites, heart behavior, ECG

Procedia PDF Downloads 85

Authors: Fotis I. Giasemis, Alexandros Sopasakis

Abstract:

This study focuses on detecting traffic anomalies using generative models applied to real-time observations. By integrating a Graph Neural Network with an attention-based mechanism within the Spatiotemporal Generative Adversarial Network framework, we enhance the capture of both spatial and temporal dependencies in traffic data. Leveraging minute-by-minute observations from cameras distributed across Gothenburg, our approach provides a more detailed and precise anomaly detection system, effectively capturing the complex topology and dynamics of urban traffic networks.

Keywords: traffic, anomaly detection, GNN, GAN

Procedia PDF Downloads 8

Authors: Mouli Banga, Chaya Ravindra

Abstract:

Wound classification is a crucial step in wound diagnosis. An effective classifier can aid wound specialists in identifying wound types with reduced financial and time investments, facilitating the determination of optimal treatment procedures. This study presents a deep neural network-based classifier that leverages wound images and their corresponding locations to categorize wounds into various classes, such as diabetic, pressure, surgical, and venous ulcers. By incorporating a developed body map, the process of tagging wound locations is significantly enhanced, providing healthcare specialists with a more efficient tool for wound analysis. We conducted a comparative analysis between two prominent convolutional neural network models, ResNet50 and MobileNetV2, utilizing a dataset of 730 images. Our findings reveal that the RestNet50 outperforms MovileNetV2, achieving an accuracy of approximately 90%, compared to MobileNetV2’s 83%. This disparity highlights the superior capability of ResNet50 in the context of this dataset. The results underscore the potential of integrating deep learning with spatial data to improve the precision and efficiency of wound diagnosis, ultimately contributing to better patient outcomes and reducing healthcare costs.

Keywords: wound classification, MobileNetV2, ResNet50, multimodel

Procedia PDF Downloads 32

Authors: Jianhong Xiang, Rui Sun, Linyu Wang

Abstract:

In recent years, due to the continuous improvement of deep learning theory, Convolutional Neural Network (CNN) has played a great superior performance in the research of Hyperspectral Image (HSI) classification. Since HSI has rich spatial-spectral information, only utilizing a single dimensional or single size convolutional kernel will limit the detailed feature information received by CNN, which limits the classification accuracy of HSI. In this paper, we design a multi-scale CNN with MLP based on residual and attention modules (INRAM-3DCNN) for the HSI classification task. We propose to use multiple 3D convolutional kernels to extract the packet feature information and fully learn the spatial-spectral features of HSI while designing residual 3D convolutional branches to avoid the decline of classification accuracy due to network degradation. Secondly, we also design the 2D Inception module with a joint channel attention mechanism to quickly extract key spatial feature information at different scales of HSI and reduce the complexity of the 3D model. Due to the high parallel processing capability and nonlinear global action of the Multilayer Perceptron (MLP), we use it in combination with the previous CNN structure for the final classification process. The experimental results on two HSI datasets show that the proposed INRAM-3DCNN method has superior classification performance and can perform the classification task excellently.

Keywords: INRAM-3DCNN, residual, channel attention, hyperspectral image classification

Procedia PDF Downloads 79

Authors: Young Kwon Yang, Bo Rang Park, Hyo Eun Lee, Tea Won Kim, Eun Ji Choi, Jin Chul Park

Abstract:

To develop an artificial intelligence based automatic ventilation system, recent research trends were analyzed and analyzed. This research method is as follows. In the field of architecture and window technology, the use of artificial intelligence, the existing study of machine learning model and the theoretical review of the literature were carried out. This paper collected journals such as Journal of Energy and Buildings, Journal of Renewable and Sustainable Energy Reviews, and articles published on Web-sites. The following keywords were searched for articles from 2000 to 2016. We searched for the above keywords mainly in the title, keyword, and abstract. As a result, the global artificial intelligence market is expected to grow at a CAGR of 14.0% from USD127bn in 2015 to USD165bn in 2017. Start-up investments in artificial intelligence increased from the US $ 45 million in 2010 to the US $ 310 million in 2015, and the number of investments increased from 6 to 54. Although AI is making efforts to advance to advanced countries, the level of technology is still in its infant stage. Especially in the field of architecture, artificial intelligence (AI) is very rare. Based on the data of this study, it is expected that the application of artificial intelligence and the application of architectural field will be revitalized through the activation of artificial intelligence in the field of architecture and window.

Keywords: artificial intelligence, window, fine dust, thermal comfort, ventilation system

Procedia PDF Downloads 275

Authors: Monireh Ahmadi Bani, Adel Khorramrouz, Lalenoor Morvarid, Bagheri Mahtab

Abstract:

Background:- Foot disorders are common in musculoskeletal problems. Plantar pressure distribution measurement is one the most important part of foot disorders diagnosis for quantitative analysis. However, the association of plantar pressure and foot disorders is not clear. With the growth of dataset and machine learning methods, the relationship between foot disorders and plantar pressures can be detected. Significance of the study:- The purpose of this study was to predict the probability of common foot disorders based on peak plantar pressure distribution and center of pressure during walking. Methodologies:- 2323 participants were assessed in a foot therapy clinic between 2015 and 2021. Foot disorders were diagnosed by an experienced physician and then they were asked to walk on a force plate scanner. After the data preprocessing, due to the difference in walking time and foot size, we normalized the samples based on time and foot size. Some of force plate variables were selected as input to a deep neural network (DNN), and the probability of any each foot disorder was measured. In next step, we used support vector machine (SVM) and run dataset for each foot disorder (classification of yes or no). We compared DNN and SVM for foot disorders prediction based on plantar pressure distributions and center of pressure. Findings:- The results demonstrated that the accuracy of deep learning architecture is sufficient for most clinical and research applications in the study population. In addition, the SVM approach has more accuracy for predictions, enabling applications for foot disorders diagnosis. The detection accuracy was 71% by the deep learning algorithm and 78% by the SVM algorithm. Moreover, when we worked with peak plantar pressure distribution, it was more accurate than center of pressure dataset. Conclusion:- Both algorithms- deep learning and SVM will help therapist and patients to improve the data pool and enhance foot disorders prediction with less expense and error after removing some restrictions properly.

Keywords: deep neural network, foot disorder, plantar pressure, support vector machine

Procedia PDF Downloads 358

Authors: R. Ramesh, K. K. Shivaraman

Abstract:

The main aim of this paper is to handle the energy requirement in an efficient manner by merging the advanced digital communication and control technologies for smart grid applications. In order to reduce user home load during peak load hours, utility applies several incentives such as real-time pricing, time of use, demand response for residential customer through smart meter. However, this method provides inconvenience in the sense that user needs to respond manually to prices that vary in real time. To overcome these inconvenience, this paper proposes a convolutional neural network (CNN) with k-means clustering machine learning model which have ability to forecast energy requirement in short term, i.e., hour of the day or day of the week. By integrating our proposed technique with home energy management based on Bluetooth low energy provides predicted value to user for scheduling appliance in advanced. This paper describes detail about CNN configuration and k-means clustering algorithm for short-term energy forecasting.

Keywords: convolutional neural network, fuzzy logic, k-means clustering approach, smart home energy management

Procedia PDF Downloads 305

Authors: Ayu Nabila Kusuma Pradana, Aprinaldi Jasa Mantau, Tomohiko Akahoshi

Abstract:

The incidence of Disseminated intravascular coagulation (DIC) following gastrointestinal surgery has a poor prognosis. Therefore, it is important to determine the factors that can predict the prognosis of DIC. This study will investigate the factors that may influence the outcome of DIC in patients after gastrointestinal surgery. Eighty-one patients were admitted to the intensive care unit after gastrointestinal surgery in Kyushu University Hospital from 2003 to 2021. Acute DIC scores were estimated using the new Japanese Association for Acute Medicine (JAAM) score from before and after surgery from day 1, day 3, and day 7. Acute DIC scores will be compared with The Sequential Organ Failure Assessment (SOFA) score, platelet count, lactate level, and a variety of biochemical parameters. This study applied machine learning algorithms to predict the prognosis of DIC after gastrointestinal surgery. The results of this study are expected to be used as an indicator for evaluating patient prognosis so that it can increase life expectancy and reduce mortality from cases of DIC patients after gastrointestinal surgery.

Keywords: the survival rate, gastrointestinal surgery, JAAM score, neural network, machine learning, disseminated intravascular coagulation (DIC)

Procedia PDF Downloads 260

Authors: Sandip Sadhukhan, Arpita Sarkar, Debprasad Sinha, Goutam Kumar Ghorai, Gautam Sarkar, Ashis K. Dhara

Abstract:

Accurate segmentation of the optic disc is very important for computer-aided diagnosis of several ocular diseases such as glaucoma, diabetic retinopathy, and hypertensive retinopathy. The paper presents an accurate and fast optic disc detection and segmentation method using an attention based fully convolutional network. The network is trained from scratch using the fundus images of extended MESSIDOR database and the trained model is used for segmentation of optic disc. The false positives are removed based on morphological operation and shape features. The result is evaluated using three-fold cross-validation on six public fundus image databases such as DIARETDB0, DIARETDB1, DRIVE, AV-INSPIRE, CHASE DB1 and MESSIDOR. The attention based fully convolutional network is robust and effective for detection and segmentation of optic disc in the images affected by diabetic retinopathy and it outperforms existing techniques.

Keywords: attention-based fully convolutional network, optic disc detection and segmentation, retinal fundus image, screening of ocular diseases

Procedia PDF Downloads 142

Authors: Yu-ding Du, Qi-lian Bao, Nassim Bessaad, Lin Liu

Abstract:

The Global Navigation Satellite System (GNSS) is regarded as an effective approach for the purpose of replacing the large amount used track-side balises in modern train localization systems. This paper describes a method based on the data fusion of a GNSS receiver sensor and an odometer sensor that can significantly improve the positioning accuracy. A digital track map is needed as another sensor to project two-dimensional GNSS position to one-dimensional along-track distance due to the fact that the train’s position can only be constrained on the track. A model trained by BP neural network is used to estimate the trend positioning error which is related to the specific location and proximate processing of the digital track map. Considering that in some conditions the satellite signal failure will lead to the increase of GNSS positioning error, a detection step for GNSS signal is applied. An adaptive weighted fusion algorithm is presented to reduce the standard deviation of train speed measurement. Finally an Extended Kalman Filter (EKF) is used for the fusion of the projected 1-D GNSS positioning data and the 1-D train speed data to get the estimate position. Experimental results suggest that the proposed method performs well, which can reduce positioning error notably.

Keywords: multi-sensor data fusion, train positioning, GNSS, odometer, digital track map, map matching, BP neural network, adaptive weighted fusion, Kalman filter

Procedia PDF Downloads 252

Authors: Natasha Petrovska, Mirjana Pavlovic, Maria M. Larrondo-Petrie

Abstract:

Neural stem cells (NSCs) are multi-potent, self-renewing cells that generate new neurons. Three subtypes of NSCs can be separated regarding the stages of NSC lineage: quiescent neural stem cells (qNSCs), activated neural stem cells (aNSCs) and neural progenitor cells (NPCs), but their gene expression signatures are not utterly understood yet. Single-cell examinations have started to elucidate the complex structure of NSC populations. Nevertheless, there is a lack of thorough molecular interpretation of the NSC lineage heterogeneity and an increasing need for tools to analyze and improve the efficiency and correctness of single-cell sequencing data. Feature selection and ordering can identify and classify the gene expression signatures of these subtypes and can discover novel subpopulations during the NSCs activation and differentiation processes. The aim here is to review the implementation of the feature selection technique on NSC subtypes and the classification techniques that have been used for the identification of gene expression signatures.

Keywords: feature selection, feature similarity, neural stem cells, genes, feature selection methods

Procedia PDF Downloads 152

Authors: Paula C. Latorre Arroyo, Andrea C. Latorre Arroyo

Abstract:

In the present society, children are exposed to and impacted by technology from very early on in various ways. Artificial intelligence (AI), in particular, directly affects them, be it positively or negatively. The concept of artificial intelligence is commonly defined as the technological programming of computers or robotic mechanisms with humanlike capabilities and characteristics. These technologies are often designed as helpful machines or disguised as handy tools that could ultimately steal private information for illicit purposes. Children, being one of the most vulnerable groups due to their lack of experience and knowledge, do not have the ability to recognize or have the malice to distinguish if an apparatus with artificial intelligence is good or bad for them. For this reason, as a society, there must be a sense of responsibility to regulate and monitor different types of uses for artificial intelligence to protect children from potential risks that might arise. This article aims to investigate the many implications that artificial intelligence has in the lives of children, starting from a home setting, within the classroom, and, ultimately, in online spaces. Irrefutably, there are numerous beneficial aspects to the use of artificial intelligence. However, due to its limitless potential and lack of specific and substantial regulations to prevent the illicit use of this technology, safety and privacy concerns surface, specifically regarding the youth. This written work aims to provide an in-depth analysis of how artificial intelligence can both help children and jeopardize their safety. Concluding with resources and data supporting the aforementioned statement.

Keywords: artificial intelligence, children, privacy, rights, safety

Procedia PDF Downloads 66

Authors: Reza Safdari, Goli Arji, Robab Abdolkhani Maryam zahmatkeshan

Abstract:

Background and purpose: Cardiovascular diseases are among the most common diseases in all societies. The most important step in minimizing myocardial infarction and its complications is to minimize its risk factors. The amount of medical data is increasingly growing. Medical data mining has a great potential for transforming these data into information. Using data mining techniques to generate predictive models for identifying those at risk for reducing the effects of the disease is very helpful. The present study aimed to collect data related to risk factors of heart infarction from patients’ medical record and developed predicting models using data mining algorithm. Methods: The present work was an analytical study conducted on a database containing 350 records. Data were related to patients admitted to Shahid Rajaei specialized cardiovascular hospital, Iran, in 2011. Data were collected using a four-sectioned data collection form. Data analysis was performed using SPSS and Clementine version 12. Seven predictive algorithms and one algorithm-based model for predicting association rules were applied to the data. Accuracy, precision, sensitivity, specificity, as well as positive and negative predictive values were determined and the final model was obtained. Results: five parameters, including hypertension, DLP, tobacco smoking, diabetes, and A+ blood group, were the most critical risk factors of myocardial infarction. Among the models, the neural network model was found to have the highest sensitivity, indicating its ability to successfully diagnose the disease. Conclusion: Risk prediction models have great potentials in facilitating the management of a patient with a specific disease. Therefore, health interventions or change in their life style can be conducted based on these models for improving the health conditions of the individuals at risk.

Keywords: decision trees, neural network, myocardial infarction, Data Mining

Procedia PDF Downloads 429

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 73

Authors: Sara ElElimy, Samir Moustafa

Abstract:

Mobile network operators start to face many challenges in the digital era, especially with high demands from customers. Since mobile network operators are considered a source of big data, traditional techniques are not effective with new era of big data, Internet of things (IoT) and 5G; as a result, handling effectively different big datasets becomes a vital task for operators with the continuous growth of data and moving from long term evolution (LTE) to 5G. So, there is an urgent need for effective Big data analytics to predict future demands, traffic, and network performance to full fill the requirements of the fifth generation of mobile network technology. In this paper, we introduce data science techniques using machine learning and deep learning algorithms: the autoregressive integrated moving average (ARIMA), Bayesian-based curve fitting, and recurrent neural network (RNN) are employed for a data-driven application to mobile network operators. The main framework included in models are identification parameters of each model, estimation, prediction, and final data-driven application of this prediction from business and network performance applications. These models are applied to Telecom Italia Big Data challenge call detail records (CDRs) datasets. The performance of these models is found out using a specific well-known evaluation criteria shows that ARIMA (machine learning-based model) is more accurate as a predictive model in such a dataset than the RNN (deep learning model).

Keywords: big data analytics, machine learning, CDRs, 5G

Procedia PDF Downloads 139

Authors: Jaiden Xuan Schraut, Leon Liu, Yiqiao Yin

Abstract:

Computer vision in medical diagnosis has achieved a high level of success in diagnosing diseases with high accuracy. However, conventional classifiers that produce an image to-label result provides insufficient information for medical professionals to judge and raise concerns over the trust and reliability of a model with results that cannot be explained. In order to gain local insight into cancerous regions, separate tasks such as imaging segmentation need to be implemented to aid the doctors in treating patients, which doubles the training time and costs which renders the diagnosis system inefficient and difficult to be accepted by the public. To tackle this issue and drive AI-first medical solutions further, this paper proposes a multi-output network that follows a U-Net architecture for image segmentation output and features an additional convolutional neural networks (CNN) module for auxiliary classification output. Class activation maps are a method of providing insight into a convolutional neural network’s feature maps that leads to its classification but in the case of lung diseases, the region of interest is enhanced by U-net-assisted Class Activation Map (CAM) visualization. Therefore, our proposed model combines image segmentation models and classifiers to crop out only the lung region of a chest X-ray’s class activation map to provide a visualization that improves the explainability and is able to generate classification results simultaneously which builds trust for AI-led diagnosis systems. The proposed U-Net model achieves 97.61% accuracy and a dice coefficient of 0.97 on testing data from the COVID-QU-Ex Dataset which includes both diseased and healthy lungs.

Keywords: multi-output network model, U-net, class activation map, image classification, medical imaging analysis

Procedia PDF Downloads 203

Authors: Kuang-Hsiung Tan, Chih-Chan Hu, Chien-Wu Lan, Shih-Sung Lin, Te-Jen Chang

Abstract:

An active islanding detection method using disturbance signal injection with intelligent controller is proposed in this study. First, a DC\AC power inverter is emulated in the distributed generator (DG) system to implement the tracking control of active power, reactive power outputs and the islanding detection. The proposed active islanding detection method is based on injecting a disturbance signal into the power inverter system through the d-axis current which leads to a frequency deviation at the terminal of the RLC load when the utility power is disconnected. Moreover, in order to improve the transient and steady-state responses of the active power and reactive power outputs of the power inverter, and to further improve the performance of the islanding detection method, two probabilistic fuzzy neural networks (PFNN) are adopted to replace the traditional proportional-integral (PI) controllers for the tracking control and the islanding detection. Furthermore, the network structure and the online learning algorithm of the PFNN are introduced in detail. Finally, the feasibility and effectiveness of the tracking control and the proposed active islanding detection method are verified with experimental results.

Keywords: distributed generators, probabilistic fuzzy neural network, islanding detection, non-detection zone

Procedia PDF Downloads 389

Authors: Diego Aguirre, Olac Fuentes

Abstract:

In this paper, we present a taxonomy of weight initialization schemes used in deep learning. We survey the most representative techniques in each class and compare them in terms of overhead cost, convergence rate, and applicability. We also introduce a new weight initialization scheme. In this technique, we perform an initial feedforward pass through the network using an initialization mini-batch. Using statistics obtained from this pass, we initialize the weights of the network, so the following properties are met: 1) weight matrices are orthogonal; 2) ReLU layers produce a predetermined number of non-zero activations; 3) the output produced by each internal layer has a unit variance; 4) weights in the last layer are chosen to minimize the error in the initial mini-batch. We evaluate our method on three popular architectures, and a faster converge rates are achieved on the MNIST, CIFAR-10/100, and ImageNet datasets when compared to state-of-the-art initialization techniques.

Keywords: deep learning, image classification, supervised learning, weight initialization

Procedia PDF Downloads 135

Authors: Nihed Bahria El Asghar, Imen Jouili, Mounir Frikha

Abstract:

Inter-datacenter transport network is very bandwidth and delay demanding. The data transferred over such a network is also highly QoS-exigent mostly because a huge volume of data should be transported transparently with regard to the application user. To avoid the data transfer failure, a backup path should be reserved. No re-routing delay should be observed. A dedicated 1+1 protection is however not applicable in inter-datacenter transport network because of the huge spare capacity. In this context, we propose a survivable virtual network with minimal backup based on network coding (1 ⊕ 1) and solve it using a modified Dijkstra-based heuristic.

Keywords: network coding, dedicated protection, spare capacity, inter-datacenters transport network

Procedia PDF Downloads 447

Authors: Jianghao Hu, Hongyu Wang

Abstract:

The effective and efficient human pose estimation algorithm is an important task for real-time human pose estimation on mobile devices. This paper proposes a light-weight human key points detection algorithm, Light-Weight Network for Real-Time Pose Estimation (LWPE). LWPE uses light-weight backbone network and depthwise separable convolutions to reduce parameters and lower latency. LWPE uses the feature pyramid network (FPN) to fuse the high-resolution, semantically weak features with the low-resolution, semantically strong features. In the meantime, with multi-scale prediction, the predicted result by the low-resolution feature map is stacked to the adjacent higher-resolution feature map to intermediately monitor the network and continuously refine the results. At the last step, the key point coordinates predicted in the highest-resolution are used as the final output of the network. For the key-points that are difficult to predict, LWPE adopts the online hard key points mining strategy to focus on the key points that hard predicting. The proposed algorithm achieves excellent performance in the single-person dataset selected in the AI (artificial intelligence) challenge dataset. The algorithm maintains high-precision performance even though the model only contains 3.9M parameters, and it can run at 225 frames per second (FPS) on the generic graphics processing unit (GPU).

Keywords: depthwise separable convolutions, feature pyramid network, human pose estimation, light-weight backbone

Procedia PDF Downloads 154

Authors: Yuanjin Liu

Abstract:

Retirement withdrawal strategies are very important to minimize the probability of ruin in retirement. The ruin probability is modeled as a function of initial withdrawal age, gender, asset allocation, inflation rate, and initial withdrawal rate. The ruin probability is obtained based on the 2019 period life table for the Social Security, IRS Required Minimum Distribution (RMD) Worksheets, US historical bond and equity returns, and inflation rates using simulation. Several popular machine learning algorithms of the generalized additive model, random forest, support vector machine, extreme gradient boosting, and artificial neural network are built. The model validation and selection are based on the test errors using hyperparameter tuning and train-test split. The optimal model is recommended for retirees to monitor the ruin probability. The optimal withdrawal strategy can be obtained based on the optimal predictive model.

Keywords: ruin probability, retirement withdrawal strategies, predictive models, optimal model

Procedia PDF Downloads 74

Authors: Tiago do Carmo Nogueira, Cássio Dener Noronha Vinhal, Gélson da Cruz Júnior, Matheus Rudolfo Diedrich Ullmann

Abstract:

Generating automatic image descriptions through natural language is a challenging task. Image captioning is a task that consistently describes an image by combining computer vision and natural language processing techniques. To accomplish this task, cutting-edge models use encoder-decoder structures. Thus, Convolutional Neural Networks (CNN) are used to extract the characteristics of the images, and Recurrent Neural Networks (RNN) generate the descriptive sentences of the images. However, cutting-edge approaches still suffer from problems of generating incorrect captions and accumulating errors in the decoders. To solve this problem, we propose a model based on the encoder-decoder structure, introducing a module that generates the weights according to the importance of the word to form the sentence, using the part-of-speech (PoS). Thus, the results demonstrate that our model surpasses state-of-the-art models.

Keywords: gated recurrent units, caption generation, convolutional neural network, part-of-speech

Procedia PDF Downloads 102

Authors: Luis C. Parra

Abstract:

The significant wave height prediction is an issue of great interest in the field of coastal activities because of the non-linear behavior of the wave height and its complexity of prediction. This study aims to present a machine learning model to forecast the significant wave height of the oceanographic wave measuring buoys anchored at Mooloolaba of the Queensland Government Data. Modeling was performed by a multilayer perceptron neural network-genetic algorithm (GA-MLP), considering Relu(x) as the activation function of the MLPNN. The GA is in charge of optimized the MLPNN hyperparameters (learning rate, hidden layers, neurons, and activation functions) and wrapper feature selection for the window width size. Results are assessed using Mean Square Error (MSE), Root Mean Square Error (RMSE), and Mean Absolute Error (MAE). The GAMLPNN algorithm was performed with a population size of thirty individuals for eight generations for the prediction optimization of 5 steps forward, obtaining a performance evaluation of 0.00104 MSE, 0.03222 RMSE, 0.02338 MAE, and 0.71163% of MAPE. The results of the analysis suggest that the MLPNNGA model is effective in predicting significant wave height in a one-step forecast with distant time windows, presenting 0.00014 MSE, 0.01180 RMSE, 0.00912 MAE, and 0.52500% of MAPE with 0.99940 of correlation factor. The GA-MLP algorithm was compared with the ARIMA forecasting model, presenting better performance criteria in all performance criteria, validating the potential of this algorithm.

Keywords: significant wave height, machine learning optimization, multilayer perceptron neural networks, evolutionary algorithms

Procedia PDF Downloads 107