Search results for: convolutional neural net

Authors: Leah Wright, Trevor Townsend

Abstract:

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

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

Procedia PDF Downloads 389

Authors: Mrinmoy Majumder, Apu Kumar Saha

Abstract:

The ever-increasing human population and its demand for energy is placing stress upon conventional energy sources; and as demand for power continues to outstrip supply, the need to optimize energy distribution and utilization is emerging as an important focus for various stakeholders. The distribution of available energy must be achieved in such a way that the needs of the consumer are satisfied. However, if the availability of resources is not sufficient to satisfy consumer demand, it is necessary to find a method to select consumers based on factors such as their socio-economic or environmental impacts. Weighting consumer types in this way can help separate them based on their relative importance, and cognitive optimization of the allocation process can then be carried out so that, even on days of particularly scarce supply, the socio-economic impacts of not satisfying the needs of consumers can be minimized. In this context, the present study utilized fuzzy logic to assign weightage to different types of consumers based at an educational campus in India, and then established optimal allocation by applying the non-linear mapping capability of neuro-genetic algorithms. The outputs of the algorithms were compared with similar outputs from particle swarm optimization and differential evolution algorithms. The results of the study demonstrate an option for the optimal utilization of available energy based on the socio-economic importance of consumers.

Keywords: power allocation, optimization problem, neural networks, environmental and ecological engineering

Procedia PDF Downloads 480

Authors: Rima Hleiss, Elie Bitar, Mahmoud Hassan, Mohamad Khalil

Abstract:

A comprehensive study of object recognition in the human brain requires combining both spatial and temporal analysis of brain activity. Here, we are mainly interested in three issues: the time perception of visual objects, the ability of discrimination between two particular categories (objects vs. animals), and the possibility to identify a particular spatial representation of visual objects. Our experiment consisted of acquiring dense electroencephalographic (EEG) signals during a picture-naming task comprising a set of objects and animals’ images. These EEG responses were recorded from nine participants. In order to determine the time perception of the presented visual stimulus, we analyzed the Event Related Potentials (ERPs) derived from the recorded EEG signals. The analysis of these signals showed that the brain perceives animals and objects with different time instants. Concerning the discrimination of the two categories, the support vector machine (SVM) was applied on the instantaneous EEG (excellent temporal resolution: on the order of millisecond) to categorize the visual stimuli into two different classes. The spatial differences between the evoked responses of the two categories were also investigated. The results showed a variation of the neural activity with the properties of the visual input. Results showed also the existence of a spatial pattern of electrodes over particular regions of the scalp in correspondence to their responses to the visual inputs.

Keywords: brain activity, categorization, dense EEG, evoked responses, spatio-temporal analysis, SVM, time perception

Procedia PDF Downloads 423

Authors: Mehmet Erdi Korkmaz, Mustafa Günay

Abstract:

Nowadays, grinding is frequently replaced with hard turning for reducing set up time and higher accuracy. This paper focused on mathematical modeling of average surface roughness (Ra) in hard turning of AISI L2 grade (DIN 1.2210) cold work tool steel with ceramic tools. The steel was hardened to 60±1 HRC after the heat treatment process. Cutting speed, feed rate, depth of cut and tool nose radius was chosen as the cutting conditions. The uncoated ceramic cutting tools were used in the machining experiments. The machining experiments were performed according to Taguchi L27 orthogonal array on CNC lathe. Ra values were calculated by averaging three roughness values obtained from three different points of machined surface. The influences of cutting conditions on surface roughness were evaluated as statistical and experimental. The analysis of variance (ANOVA) with 95% confidence level was applied for statistical analysis of experimental results. Finally, mathematical models were developed using the artificial neural networks (ANN). ANOVA results show that feed rate is the dominant factor affecting surface roughness, followed by tool nose radius and cutting speed.

Keywords: ANN, hard turning, DIN 1.2210, surface roughness, Taguchi method

Procedia PDF Downloads 372

Authors: Zhaohan Xie, Yining Yu, Mingliang Chen

Abstract:

As research has shown a convergent tendency for aversion to AI recommendation, it is imperative to find a way to promote AI usage and better harness the technology. In the context of e-commerce, this study has found evidence that people show less avoidance of algorithms when recommending search products compared to experience products. This is due to people’s different attribution of mind to AI versus humans, as suggested by mind perception theory. While people hold the belief that an algorithm owns sufficient capability to think and calculate, which makes it competent to evaluate search product attributes that can be obtained before actual use, they doubt its capability to sense and feel, which is essential for evaluating experience product attributes that must be assessed after experience in person. The result of the behavioral investigation (Study 1, N=112) validated that consumers show low purchase intention to experience products recommended by AI. Further consumer neuroscience study (Study 2, N=26) using Event-related potential (ERP) showed that consumers have a higher level of cognitive conflict when faced with AI recommended experience product as reflected by larger N2 component, while the effect disappears for search product. This research has implications for the effective employment of AI recommenders, and it extends the literature on e-commerce and marketing communication.

Keywords: algorithm recommendation, consumer behavior, e-commerce, event-related potential, experience product, search product

Procedia PDF Downloads 157

Authors: Mohamed Abdelrahman Abdalla

Abstract:

Apart from their function in producing CSF, the brain ventricles have been recognized as the mere remnant of the embryological neural tube with no clear role. The lack of proper definition of the function of the brain ventricles and the central spinal canal has made it difficult to ascertain the pathophysiology of its different disease conditions or to treat them. This study aims to review the simple physics that could explain the basic function of the CNS ventricular system and to suggest new ways of approaching its pathology. There are probably more physical factors to consider than only the pressure. Monro-Killie hypothesis focuses on volume and subsequently pressure to direct our surgical management in different disease conditions. However, the enlarged volume of the ventricles in normal pressure hydrocephalus does not move any blood or brain outside the skull. Also, in idiopathic intracranial hypertension, the very high intracranial pressure rarely causes brain herniation. On this note, the continuum of the intracranial cavity with the spinal canal makes it a whole unit and hence the defect in the theory. In this study, adding different factors to the equation like brain and CSF density and positions of the brain in space, in addition to the volume and pressure, aims to identify how the ventricles are important in the CNS homeostasis. In addition, increasing the variables that we analyze to treat different CSF pathological conditions should increase our understanding and hence accuracy of treatment of such conditions.

Keywords: communicating hydrocephalus, functions of the ventricles, idiopathic intracranial hypertension physics of CSF

Procedia PDF Downloads 108

Authors: Qais M. Yousef, Yasmeen A. Alshaer

Abstract:

Over the last few years, the amount of data available on the globe has been increased rapidly. This came up with the emergence of recent concepts, such as the big data and the Internet of Things, which have furnished a suitable solution for the availability of data all over the world. However, managing this massive amount of data remains a challenge due to their large verity of types and distribution. Therefore, locating the required file particularly from the first trial turned to be a not easy task, due to the large similarities of names for different files distributed on the web. Consequently, the accuracy and speed of search have been negatively affected. This work presents a method using Electroencephalography signals to locate the files based on their contents. Giving the concept of natural mind waves processing, this work analyses the mind wave signals of different people, analyzing them and extracting their most appropriate features using multi-objective metaheuristic algorithm, and then classifying them using artificial neural network to distinguish among files with similar names. The aim of this work is to provide the ability to find the files based on their contents using human thoughts only. Implementing this approach and testing it on real people proved its ability to find the desired files accurately within noticeably shorter time and retrieve them as a first choice for the user.

Keywords: artificial intelligence, data contents search, human active memory, mind wave, multi-objective optimization

Procedia PDF Downloads 179

Authors: Daneal Rorke, Gueguim Kana

Abstract:

The shift towards renewable energy sources for biofuel production has received increasing attention. However, the use and pre-treatment of lignocellulosic material are inundated with the generation of fermentation inhibitors which severely impact the feasibility of bioprocesses. This study reports the profiling of all volatile compounds generated during microwave assisted chemical pre-treatment of sorghum leaves. Furthermore, the optimization of reducing sugar (RS) from microwave assisted acid pre-treatment of sorghum leaves was assessed and gave a coefficient of determination (R2) of 0.76, producing an optimal RS yield of 2.74 g FS/g substrate. The development of an intelligent model to predict volatile compound fractions gave R2 values of up to 0.93 for 21 volatile compounds. Sensitivity analysis revealed that furfural and phenol exhibited high sensitivity to acid concentration, alkali concentration and S:L ratio, while phenol showed high sensitivity to microwave duration and intensity as well. These findings illustrate the potential of using an intelligent model to predict the volatile compound fraction profile of compounds generated during pre-treatment of sorghum leaves in order to establish a more robust and efficient pre-treatment regime for biofuel production.

Keywords: artificial neural networks, fermentation inhibitors, lignocellulosic pre-treatment, sorghum leaves

Procedia PDF Downloads 249

Authors: David Jurado, Carlos Ávila

Abstract:

Detection of early signs of breast cancer development is crucial to quickly diagnose the disease and to define adequate treatment to increase the survival probability of the patient. Computer Aided Detection systems (CADs), along with modern data techniques such as Machine Learning (ML) and Neural Networks (NN), have shown an overall improvement in digital mammography cancer diagnosis, reducing the false positive and false negative rates becoming important tools for the diagnostic evaluations performed by specialized radiologists. However, ML and NN-based algorithms rely on datasets that might bring issues to the segmentation tasks. In the present work, an automatic segmentation and detection algorithm is described. This algorithm uses image processing techniques along with the Hough transform to automatically identify microcalcifications that are highly correlated with breast cancer development in the early stages. Along with image processing, automatic segmentation of high-contrast objects is done using edge extraction and circle Hough transform. This provides the geometrical features needed for an automatic mask design which extracts statistical features of the regions of interest. The results shown in this study prove the potential of this tool for further diagnostics and classification of mammographic images due to the low sensitivity to noisy images and low contrast mammographies.

Keywords: breast cancer, segmentation, X-ray imaging, hough transform, image analysis

Procedia PDF Downloads 84

Authors: Bi Zhao

Abstract:

Neural network technology has greatly improved the output of machine translation in terms of both fluency and accuracy, which greatly increases its appeal for young users. The present exploratory study aims to find out how the Chinese undergraduates perceive and use machine translation in their daily life. A survey is conducted to collect data from 100 undergraduate students from multiple Chinese universities and with varied academic backgrounds, including arts, business, science, engineering, and medicine. The survey questions inquire about their use (including frequency, scenarios, purposes, and preferences) of and attitudes (including trust, quality assessment, justifications, and ethics) toward machine translation. Interviews and tasks of evaluating machine translation output are also employed in combination with the survey on a sample of selected respondents. The results indicate that Chinese undergraduate students use machine translation on a daily basis for a wide range of purposes in academic, communicative, and entertainment scenarios. Most of them have preferred machine translation tools, but the availability of machine translation tools within a certain scenario, such as the embedded machine translation tool on the webpage, is also the determining factor in their choice. The results also reveal that despite the reportedly limited trust in the accuracy of machine translation output, most students lack the ability to critically analyze and evaluate such output. Furthermore, the evidence is revealed of the inadequate awareness of ethical responsibility as machine translation users among Chinese undergraduate students.

Keywords: Chinese undergraduates, machine translation, trust, usage

Procedia PDF Downloads 139

Authors: Parker Townes, Prabdeep Panesar, Chunlin Liu, Soo Youn Lee, Dan Devoe, Paul D. Arnold, Jennifer Crosbie, Russell Schachar

Abstract:

Attention-deficit hyperactivity disorder (ADHD) and autism spectrum disorder (ASD) are impairing childhood neurodevelopmental disorders with problems in executive functions. Executive functions are higher-level mental processes essential for daily functioning and goal attainment. There is genetic and neural overlap between ADHD and ASD. The aim of this meta-analysis was to evaluate if pediatric ASD and ADHD have distinct executive function profiles. This review was completed following Cochrane guidelines. Fifty-eight articles were identified through database searching, followed by a blinded screening in duplicate. A meta-analysis was performed for all task performance metrics evaluated by at least two articles. Forty-five metrics from 24 individual tasks underwent analysis. No differences were found between youth with ASD and ADHD in any domain under direct comparison. However, individuals with ASD and ADHD exhibited deficient attention, flexibility, visuospatial abilities, working memory, processing speed, and response inhibition compared to controls. No deficits in planning were noted in either disorder. Only 11 studies included a group with comorbid ASD+ADHD, making it difficult to determine whether common executive function deficits are a function of comorbidity. Further research is needed to determine if comorbidity accounts for the apparent commonality in executive function between ASD and ADHD.

Keywords: autism spectrum disorder, ADHD, neurocognition, executive function, youth

Procedia PDF Downloads 77

Authors: Bingchun Liu, Pei-Chann Chang, Natasha Huang, Dun Li

Abstract:

Machine Learning and Data Mining are the two important tools for extracting useful information and knowledge from large datasets. In machine learning, classification is a wildly used technique to predict qualitative variables and is generally preferred over regression from an operational point of view. Due to the enormous increase in air pollution in various countries especially China, Air Quality Classification has become one of the most important topics in air quality research and modelling. This study aims at introducing a hybrid classification model based on information theory and Support Vector Machine (SVM) using the air quality data of four cities in China namely Beijing, Guangzhou, Shanghai and Tianjin from Jan 1, 2014 to April 30, 2016. China's Ministry of Environmental Protection has classified the daily air quality into 6 levels namely Serious Pollution, Severe Pollution, Moderate Pollution, Light Pollution, Good and Excellent based on their respective Air Quality Index (AQI) values. Using the information theory, information gain (IG) is calculated and feature selection is done for both categorical features and continuous numeric features. Then SVM Machine Learning algorithm is implemented on the selected features with cross-validation. The final evaluation reveals that the IG and SVM hybrid model performs better than SVM (alone), Artificial Neural Network (ANN) and K-Nearest Neighbours (KNN) models in terms of accuracy as well as complexity.

Keywords: machine learning, air quality classification, air quality index, information gain, support vector machine, cross-validation

Procedia PDF Downloads 236

Authors: Andrey Khalov

Abstract:

The rapid proliferation of unstructured data in IT infrastructure management demands innovative approaches for extracting actionable knowledge. This paper presents a framework for ontology-based knowledge extraction that combines relational graph neural networks (R-GNN) with large language models (LLMs). The proposed method leverages the DOLCE framework as the foundational ontology, extending it with concepts from ITSMO for domain-specific applications in IT service management and outsourcing. A key component of this research is the use of transformer-based models, such as DeBERTa-v3-large, for automatic entity and relationship extraction from unstructured texts. Furthermore, the paper explores how transfer learning techniques can be applied to fine-tune large language models (LLaMA) for using to generate synthetic datasets to improve precision in BERT-based entity recognition and ontology alignment. The resulting IT Ontology (ITO) serves as a comprehensive knowledge base that integrates domain-specific insights from ITIL processes, enabling more efficient decision-making. Experimental results demonstrate significant improvements in knowledge extraction and relationship mapping, offering a cutting-edge solution for enhancing cognitive computing in IT service environments.

Keywords: ontology expansion, synthetic dataset, transformer fine-tuning, concept extraction, DOLCE, BERT, taxonomy, LLM, NER

Procedia PDF Downloads 18

Authors: Nirupama Dixit, Anyaa Mittal, Neeru Sood

Abstract:

Alzheimer’s disease (AD) is a progressive form of dementia, contributing to up to 70% of cases, mostly observed in elderly but is not restricted to old age. The pathophysiology of the disease is characterized by specific pathological changes in brain. The changes (i.e. accumulation of metal ions in brain, formation of extracellular β-amyloid (Aβ) peptide aggregates and tangle of hyper phosphorylated Tau protein inside neurons) damage the neuronal connections irreversibly. The current issues in improvement of life quality of Alzheimer's patient lies in the fact that the diagnosis is made at a late stage of the disease and the medications do not treat the basic causes of Alzheimer's. The targeted delivery of drug through the blood brain barrier (BBB) poses several limitations via traditional approaches for treatment. To overcome these drug delivery limitation, nanoparticles provide a promising solution. This review focuses on current strategies for efficient targeted drug delivery using nanoparticles and improving the quality of therapy provided to the patient. Nanoparticles can be used to encapsulate drug (which is generally hydrophobic) to ensure its passage to brain; they can be conjugated to metal ion chelators to reduce the metal load in neural tissue thus lowering the harmful effects of oxidative damage; can be conjugated with drug and monoclonal antibodies against BBB endogenous receptors. Finally this review covers how the nanoparticles can play a role in diagnosing the disease.

Keywords: Alzheimer's disease, β-amyloid plaques, blood brain barrier, metal chelators, nanoparticles

Procedia PDF Downloads 492

Authors: Parsa Mahdavi, M. Amin Hariri-Ardebili

Abstract:

The development of accurate machine learning and deep learning models traditionally demands hands-on expertise and a solid background to fine-tune hyperparameters. With the continuous expansion of datasets in various scientific and engineering domains, researchers increasingly turn to machine learning methods to unveil hidden insights that may elude classic regression techniques. This surge in adoption raises concerns about the adequacy of the resultant meta-models and, consequently, the interpretation of the findings. In response to these challenges, automated machine learning (AutoML) emerges as a promising solution, aiming to construct machine learning models with minimal intervention or guidance from human experts. AutoML encompasses crucial stages such as data preparation, feature engineering, hyperparameter optimization, and neural architecture search. This paper provides a comprehensive overview of the principles underpinning AutoML, surveying several widely-used AutoML platforms. Additionally, the paper offers a glimpse into the application of AutoML on various engineering datasets. By comparing these results with those obtained through classical machine learning methods, the paper quantifies the uncertainties inherent in the application of a single ML model versus the holistic approach provided by AutoML. These examples showcase the efficacy of AutoML in extracting meaningful patterns and insights, emphasizing its potential to revolutionize the way we approach and analyze complex datasets.

Keywords: automated machine learning, uncertainty, engineering dataset, regression

Procedia PDF Downloads 62

Authors: Krishang Surapaneni

Abstract:

The goal of this project is to determine how to predict important aspects of options, including the ask price. We want to compare different machine learning models to learn the best model and the best hyperparameters for that model for this purpose and data set. Option pricing is a relatively new field, and it can be very complicated and intimidating, especially to inexperienced people, so we want to create a machine learning model that can predict important aspects of an option stock, which can aid in future research. We tested multiple different models and experimented with hyperparameter tuning, trying to find some of the best parameters for a machine-learning model. We tested three different models: a Random Forest Regressor, a linear regressor, and an MLP (multi-layer perceptron) regressor. The most important feature in this experiment is the ask price; this is what we were trying to predict. In the field of stock pricing prediction, there is a large potential for error, so we are unable to determine the accuracy of the models based on if they predict the pricing perfectly. Due to this factor, we determined the accuracy of the model by finding the average percentage difference between the predicted and actual values. We tested the accuracy of the machine learning models by comparing the actual results in the testing data and the predictions made by the models. The linear regression model performed worst, with an average percentage error of 17.46%. The MLP regressor had an average percentage error of 11.45%, and the random forest regressor had an average percentage error of 7.42%

Keywords: finance, linear regression model, machine learning model, neural network, stock price

Procedia PDF Downloads 77

Authors: Reza Mohammadzadeh

Abstract:

The major challenges in geotechnical engineering in underground spaces arise from uncertainties and different probabilities. The collection, collation, and collaboration of existing data to incorporate them in analysis and design for given prospect evaluation would be a reliable, practical problem solving method under uncertainty. Machine learning (ML) is a subfield of artificial intelligence in statistical science which applies different techniques (e.g., Regression, neural networks, support vector machines, decision trees, random forests, genetic programming, etc.) on data to automatically learn and improve from them without being explicitly programmed and make decisions and predictions. In this paper, a conceptual database schema of geotechnical risks in underground coal mining based on a cloud system architecture has been designed. A new approach of risk assessment using a three-dimensional risk matrix supported by the level of knowledge (LoK) has been proposed in this model. Subsequently, the model workflow methodology stages have been described. In order to train data and LoK models deployment, an ML platform has been implemented. IBM Watson Studio, as a leading data science tool and data-driven cloud integration ML platform, is employed in this study. As a Use case, a data set of geotechnical hazards and risk assessment in underground coal mining were prepared to demonstrate the performance of the model, and accordingly, the results have been outlined.

Keywords: data model, geotechnical risks, machine learning, underground coal mining

Procedia PDF Downloads 275

Authors: B. Lokuketagoda, M. Weerakoon, U. Madushan, A. N. Senaratne, K. Y. Abeywardena

Abstract:

Ransomware has become a common threat in past few years and the recent threat reports show an increase of growth in Ransomware infections. Researchers have identified different variants of Ransomware families since 2015. Lack of knowledge of the user about the threat is a major concern. Ransomware detection methodologies are still growing through the industry. Email is the easiest method to send Ransomware to its victims. Uninformed users tend to click on links and attachments without much consideration assuming the emails are genuine. As a solution to this in this paper R-Killer Ransomware detection tool is introduced. Tool can be integrated with existing email services. The core detection Engine (CDE) discussed in the paper focuses on separating suspicious samples from emails and handling them until a decision is made regarding the suspicious mail. It has the capability of preventing execution of identified ransomware processes. On the other hand, Sandboxing and URL analyzing system has the capability of communication with public threat intelligence services to gather known threat intelligence. The R-Killer has its own mechanism developed in its Proactive Monitoring System (PMS) which can monitor the processes created by downloaded email attachments and identify potential Ransomware activities. R-killer is capable of gathering threat intelligence without exposing the user’s data to public threat intelligence services, hence protecting the confidentiality of user data.

Keywords: ransomware, deep learning, recurrent neural networks, email, core detection engine

Procedia PDF Downloads 216

Authors: Getaneh Berie Tarekegn, Li-Chia Tai

Abstract:

With the rapid advancement of artificial intelligence, low-power built-in sensors on Internet of Things devices, and communication technologies, location-aware services have become increasingly popular and have permeated every aspect of people’s lives. Global navigation satellite systems (GNSSs) are the default method of providing continuous positioning services for ground and aerial vehicles, as well as consumer devices (smartphones, watches, notepads, etc.). However, the environment affects satellite positioning systems, particularly indoors, in dense urban and suburban cities enclosed by skyscrapers, or when deep shadows obscure satellite signals. This is because (1) indoor environments are more complicated due to the presence of many objects surrounding them; (2) reflection within the building is highly dependent on the surrounding environment, including the positions of objects and human activity; and (3) satellite signals cannot be reached in an indoor environment, and GNSS doesn't have enough power to penetrate building walls. GPS is also highly power-hungry, which poses a severe challenge for battery-powered IoT devices. Due to these challenges, IoT applications are limited. Consequently, precise, seamless, and ubiquitous Positioning, Navigation and Timing (PNT) systems are crucial for many artificial intelligence Internet of Things (AI-IoT) applications in the era of smart cities. Their applications include traffic monitoring, emergency alarms, environmental monitoring, location-based advertising, intelligent transportation, and smart health care. This paper proposes a generative AI-based positioning scheme for large-scale wireless settings using fingerprinting techniques. In this article, we presented a semi-supervised deep convolutional generative adversarial network (S-DCGAN)-based radio map construction method for real-time device localization. We also employed a reliable signal fingerprint feature extraction method with t-distributed stochastic neighbor embedding (t-SNE), which extracts dominant features while eliminating noise from hybrid WLAN and long-term evolution (LTE) fingerprints. The proposed scheme reduced the workload of site surveying required to build the fingerprint database by up to 78.5% and significantly improved positioning accuracy. The results show that the average positioning error of GAILoc is less than 0.39 m, and more than 90% of the errors are less than 0.82 m. According to numerical results, SRCLoc improves positioning performance and reduces radio map construction costs significantly compared to traditional methods.

Keywords: location-aware services, feature extraction technique, generative adversarial network, long short-term memory, support vector machine

Procedia PDF Downloads 44

Authors: Kun Xu, Yuan Xu, Jia Qiao

Abstract:

The document detection plays an important role in optical character recognition and text analysis. Because the traditional detection methods have weak generalization ability, and deep neural network has complex structure and large number of parameters, which cannot be well applied in mobile devices, this paper proposes a lightweight Document Corner Detection Network (DCDNet). DCDNet is a two-stage architecture. The first stage with Encoder-Decoder structure adopts depthwise separable convolution to greatly reduce the network parameters. After introducing the Feature Attention Union (FAU) module, the second stage enhances the feature information of spatial and channel dim and adaptively adjusts the size of receptive field to enhance the feature expression ability of the model. Aiming at solving the problem of the large difference in the number of pixel distribution between corner and non-corner, Weighted Binary Cross Entropy Loss (WBCE Loss) is proposed to define corner detection problem as a classification problem to make the training process more efficient. In order to make up for the lack of Dataset of document corner detection, a Dataset containing 6620 images named Document Corner Detection Dataset (DCDD) is made. Experimental results show that the proposed method can obtain fast, stable and accurate detection results on DCDD.

Keywords: document detection, corner detection, attention mechanism, lightweight

Procedia PDF Downloads 354

Authors: Getaneh Berie Tarekegn, Li-Chia Tai

Abstract:

With the rapid advancement of artificial intelligence, low-power built-in sensors on Internet of Things devices, and communication technologies, location-aware services have become increasingly popular and have permeated every aspect of people’s lives. Global navigation satellite systems (GNSSs) are the default method of providing continuous positioning services for ground and aerial vehicles, as well as consumer devices (smartphones, watches, notepads, etc.). However, the environment affects satellite positioning systems, particularly indoors, in dense urban and suburban cities enclosed by skyscrapers, or when deep shadows obscure satellite signals. This is because (1) indoor environments are more complicated due to the presence of many objects surrounding them; (2) reflection within the building is highly dependent on the surrounding environment, including the positions of objects and human activity; and (3) satellite signals cannot be reached in an indoor environment, and GNSS doesn't have enough power to penetrate building walls. GPS is also highly power-hungry, which poses a severe challenge for battery-powered IoT devices. Due to these challenges, IoT applications are limited. Consequently, precise, seamless, and ubiquitous Positioning, Navigation and Timing (PNT) systems are crucial for many artificial intelligence Internet of Things (AI-IoT) applications in the era of smart cities. Their applications include traffic monitoring, emergency alarming, environmental monitoring, location-based advertising, intelligent transportation, and smart health care. This paper proposes a generative AI-based positioning scheme for large-scale wireless settings using fingerprinting techniques. In this article, we presented a novel semi-supervised deep convolutional generative adversarial network (S-DCGAN)-based radio map construction method for real-time device localization. We also employed a reliable signal fingerprint feature extraction method with t-distributed stochastic neighbor embedding (t-SNE), which extracts dominant features while eliminating noise from hybrid WLAN and long-term evolution (LTE) fingerprints. The proposed scheme reduced the workload of site surveying required to build the fingerprint database by up to 78.5% and significantly improved positioning accuracy. The results show that the average positioning error of GAILoc is less than 0.39 m, and more than 90% of the errors are less than 0.82 m. According to numerical results, SRCLoc improves positioning performance and reduces radio map construction costs significantly compared to traditional methods.

Keywords: location-aware services, feature extraction technique, generative adversarial network, long short-term memory, support vector machine

Procedia PDF Downloads 50

Authors: David Percy

Abstract:

Landslides are a geologic phenomenon that affects a large number of inhabited places and are constantly being monitored and studied for the prediction of future occurrences. Reconstructability analysis (RA) is a methodology for extracting informative models from large volumes of data that work exclusively with discrete data. While RA has been used in medical applications and social science extensively, we are introducing it to the spatial sciences through applications like landslide prediction. Since RA works exclusively with discrete data, such as soil classification or bedrock type, working with continuous data, such as porosity, requires that these data are binned for inclusion in the model. RA constructs models of the data which pick out the most informative elements, independent variables (IVs), from each layer that predict the dependent variable (DV), landslide occurrence. Each layer included in the model retains its classification data as a primary encoding of the data. Unlike other machine learning algorithms that force the data into one-hot encoding type of schemes, RA works directly with the data as it is encoded, with the exception of continuous data, which must be binned. The usual physical and derived layers are included in the model, and testing our results against other published methodologies, such as neural networks, yields accuracy that is similar but with the advantage of a completely transparent model. The results of an RA session with a data set are a report on every combination of variables and their probability of landslide events occurring. In this way, every combination of informative state combinations can be examined.

Keywords: reconstructability analysis, machine learning, landslides, raster analysis

Procedia PDF Downloads 68

Authors: Farshid Hajati, Federico Girosi, Shima Ghassempour

Abstract:

We present a Deep Belief Network (DBN) method for clustering health trajectories. Deep Belief Network (DBN) is a deep architecture that consists of a stack of Restricted Boltzmann Machines (RBM). In a deep architecture, each layer learns more complex features than the past layers. The proposed method depends on DBN in clustering without using back propagation learning algorithm. The proposed DBN has a better a performance compared to the deep neural network due the initialization of the connecting weights. We use Contrastive Divergence (CD) method for training the RBMs which increases the performance of the network. The performance of the proposed method is evaluated extensively on the Health and Retirement Study (HRS) database. The University of Michigan Health and Retirement Study (HRS) is a nationally representative longitudinal study that has surveyed more than 27,000 elderly and near-elderly Americans since its inception in 1992. Participants are interviewed every two years and they collect data on physical and mental health, insurance coverage, financial status, family support systems, labor market status, and retirement planning. The dataset is publicly available and we use the RAND HRS version L, which is easy to use and cleaned up version of the data. The size of sample data set is 268 and the length of the trajectories is equal to 10. The trajectories do not stop when the patient dies and represent 10 different interviews of live patients. Compared to the state-of-the-art benchmarks, the experimental results show the effectiveness and superiority of the proposed method in clustering health trajectories.

Keywords: health trajectory, clustering, deep learning, DBN

Procedia PDF Downloads 371

Authors: Yeliz Karaca, Carlo Cattani

Abstract:

This study focuses on the local Holder exponent as a measure of the function regularity for time series related to finance data. In this study, the attributes of the finance dataset belonging to 13 countries (India, China, Japan, Sweden, France, Germany, Italy, Australia, Mexico, United Kingdom, Argentina, Brazil, USA) located in 5 different continents (Asia, Europe, Australia, North America and South America) have been examined.These countries are the ones mostly affected by the attributes with regard to financial development, covering a period from 2012 to 2017. Our study is concerned with the most important attributes that have impact on the development of finance for the countries identified. Our method is comprised of the following stages: (a) among the multi fractal methods and Brownian motion Holder regularity functions (polynomial, exponential), significant and self-similar attributes have been identified (b) The significant and self-similar attributes have been applied to the Artificial Neuronal Network (ANN) algorithms (Feed Forward Back Propagation (FFBP) and Cascade Forward Back Propagation (CFBP)) (c) the outcomes of classification accuracy have been compared concerning the attributes that have impact on the attributes which affect the countries’ financial development. This study has enabled to reveal, through the application of ANN algorithms, how the most significant attributes are identified within the relevant dataset via the Holder functions (polynomial and exponential function).

Keywords: artificial neural networks, finance data, Holder regularity, multifractals

Procedia PDF Downloads 247

Authors: S. Abera, T. Gidey, W. Terefe

Abstract:

Introduction: Testing for HIV is the key entry point to HIV prevention, treatment, and care and support services. Hence, predictive data mining techniques can greatly benefit to analyze and discover new patterns from huge datasets like that of EDHS 2011 data. Objectives: The objective of this study is to build a predictive modeling for HIV testing and identify determinants of HIV testing for adults with age above fourteen years using data mining techniques. Methods: Cross-Industry Standard Process for Data Mining (CRISP-DM) was used to predict the model for HIV testing and explore association rules between HIV testing and the selected attributes among adult Ethiopians. Decision tree, Naïve-Bayes, logistic regression and artificial neural networks of data mining techniques were used to build the predictive models. Results: The target dataset contained 30,625 study participants; of which 16, 515 (53.9%) were women. Nearly two-fifth; 17,719 (58%), have never been tested for HIV while the rest 12,906 (42%) had been tested. Ethiopians with higher wealth index, higher educational level, belonging 20 to 29 years old, having no stigmatizing attitude towards HIV positive person, urban residents, having HIV related knowledge, information about family planning on mass media and knowing a place where to get testing for HIV showed an increased patterns with respect to HIV testing. Conclusion and Recommendation: Public health interventions should consider the identified determinants to promote people to get testing for HIV.

Keywords: data mining, HIV, testing, ethiopia

Procedia PDF Downloads 499

Authors: Halefom Kidane

Abstract:

This study aims to assess the wind resource potential and characterize the urban area wind patterns in Hawassa City, Ethiopia. The estimation and characterization of wind resources are crucial for sustainable urban planning, renewable energy development, and climate change mitigation strategies. A secondary data collection method was used to carry out the study. The collected data at 2 meters was analyzed statistically and extrapolated to the standard heights of 10-meter and 30-meter heights using the power law equation. The standard deviation method was used to calculate the value of scale and shape factors. From the analysis presented, the maximum and minimum mean daily wind speed at 2 meters in 2016 was 1.33 m/s and 0.05 m/s in 2017, 1.67 m/s and 0.14 m/s in 2018, 1.61m and 0.07 m/s, respectively. The maximum monthly average wind speed of Hawassa City in 2016 at 2 meters was noticed in the month of December, which is around 0.78 m/s, while in 2017, the maximum wind speed was recorded in the month of January with a wind speed magnitude of 0.80 m/s and in 2018 June was maximum speed which is 0.76 m/s. On the other hand, October was the month with the minimum mean wind speed in all years, with a value of 0.47 m/s in 2016,0.47 in 2017 and 0.34 in 2018. The annual mean wind speed was 0.61 m/s in 2016,0.64, m/s in 2017 and 0.57 m/s in 2018 at a height of 2 meters. From extrapolation, the annual mean wind speeds for the years 2016,2017 and 2018 at 10 heights were 1.17 m/s,1.22 m/s, and 1.11 m/s, and at the height of 30 meters, were 3.34m/s,3.78 m/s, and 3.01 m/s respectively/Thus, the site consists mainly primarily classes-I of wind speed even at the extrapolated heights.

Keywords: artificial neural networks, forecasting, min-max normalization, wind speed

Procedia PDF Downloads 77

Authors: Jonas Colin

Abstract:

Standing at the vanguard of AI development, Alpha represents an unprecedented synthesis of logical rigor and human abstraction, meticulously crafted to mirror the user's unique persona and personality, a feat previously unattainable in AI development. Alpha, an avant-garde artefact in the realm of artificial intelligence, epitomizes a paradigmatic shift in personalized digital interaction, amalgamating user-specific dialogic patterns with the sophisticated algorithmic prowess of OpenAI's GPT-3.5 to engender a platform for enhanced metacognitive engagement and individualized user experience. Underpinned by a sophisticated algorithmic framework, Alpha integrates vast datasets through a complex interplay of neural network models and symbolic AI, facilitating a dynamic, adaptive learning process. This integration enables the system to construct a detailed user profile, encompassing linguistic preferences, emotional tendencies, and cognitive styles, tailoring interactions to align with individual characteristics and conversational contexts. Furthermore, Alpha incorporates advanced metacognitive elements, enabling real-time reflection and adaptation in communication strategies. This self-reflective capability ensures continuous refinement of its interaction model, positioning Alpha not just as a technological marvel but as a harbinger of a new era in human-computer interaction, where machines engage with us on a deeply personal and cognitive level, transforming our interaction with the digital world.

Keywords: chatbot, GPT 3.5, metacognition, symbiose

Procedia PDF Downloads 72

Authors: Chih-Chieh Yang, Chung-Neng Huang

Abstract:

In power system, how to improve the fault diagnosis technology of transmission line has always been the primary goal of power grid operators. In recent years, due to the rise of green energy, the addition of all kinds of distributed power also has an impact on the stability of the power system. Because the smart meters are with the function of data recording and bidirectional transmission, the adaptive Fuzzy Neural inference system, ANFIS, as well as the artificial intelligence that has the characteristics of learning and estimation in artificial intelligence. For transmission network, in order to avoid misjudgment of the fault type and location due to the input of these unstable power sources, combined with the above advantages of smart meter and ANFIS, a method for identifying fault types and location of faults is proposed in this study. In ANFIS training, the bus voltage and current information collected by smart meters can be trained through the ANFIS tool in MATLAB to generate fault codes to identify different types of faults and the location of faults. In addition, due to the uncertainty of distributed generation, a wind power system is added to the transmission network to verify the diagnosis correctness of the study. Simulation results show that the method proposed in this study can correctly identify the fault type and location of fault with more efficiency, and can deal with the interference caused by the addition of unstable power sources.

Keywords: ANFIS, fault diagnosis, power system, smart meter

Procedia PDF Downloads 140

Authors: Simone Leonardi, Luca Ardito

Abstract:

Micro-influencers are effective elements in the marketing strategies of companies and institutions because of their capability to create an hyper-engaged audience around a specific topic of interest. In recent years, many scientific approaches and commercial tools have handled the task of detecting this type of social media users. These strategies adopt solutions ranging from rule based machine learning models to deep neural networks and graph analysis on text, images, and account information. This work compares the existing solutions and proposes an ensemble method to generalize them with different input data and social media platforms. The deployed solution combines deep learning models on unstructured data with statistical machine learning models on structured data. We retrieve both social media accounts information and multimedia posts on Twitter and Instagram. These data are mapped into feature vectors for an eXtreme Gradient Boosting (XGBoost) classifier. Sixty different topics have been analyzed to build a rule based gold standard dataset and to compare the performances of our approach against baseline classifiers. We prove the effectiveness of our work by comparing the accuracy, precision, recall, and f1 score of our model with different configurations and architectures. We obtained an accuracy of 0.91 with our best performing model.

Keywords: deep learning, gradient boosting, image processing, micro-influencers, NLP, social media

Procedia PDF Downloads 184

Authors: RAM PAL SINGH, VIKASH CHAUDHARY, MONIKA VERMA

Abstract:

In this paper, we proposed the implementation of optimization based Extreme Learning Machine (ELM) for watermarking of B-channel of color image in discrete wavelet transform (DWT) domain. ELM, a regularization algorithm, works based on generalized single-hidden-layer feed-forward neural networks (SLFNs). However, hidden layer parameters, generally called feature mapping in context of ELM need not to be tuned every time. This paper shows the embedding and extraction processes of watermark with the help of ELM and results are compared with already used machine learning models for watermarking.Here, a cover image is divide into suitable numbers of non-overlapping blocks of required size and DWT is applied to each block to be transformed in low frequency sub-band domain. Basically, ELM gives a unified leaning platform with a feature mapping, that is, mapping between hidden layer and output layer of SLFNs, is tried for watermark embedding and extraction purpose in a cover image. Although ELM has widespread application right from binary classification, multiclass classification to regression and function estimation etc. Unlike SVM based algorithm which achieve suboptimal solution with high computational complexity, ELM can provide better generalization performance results with very small complexity. Efficacy of optimization method based ELM algorithm is measured by using quantitative and qualitative parameters on a watermarked image even though image is subjected to different types of geometrical and conventional attacks.

Keywords: BER, DWT, extreme leaning machine (ELM), PSNR

Procedia PDF Downloads 313