Search results for: stub tuning

Authors: Ma. Gracia Corazon Cayanan, Kai Yuen Cheong, Li Sha

Abstract:

Training a language model for a minority language has been a challenging task. The lack of available corpora to train and fine-tune state-of-the-art language models is still a challenge in the area of Natural Language Processing (NLP). Moreover, the need for high computational resources and bulk data limit the attainment of this task. In this paper, we presented the following contributions: (1) we introduce and used a translation pair set of Tagalog and English (TL-EN) in pre-training a language model to a minority language resource; (2) we fine-tuned and evaluated top-ranking and pre-trained semantic textual similarity binary task (STSB) models, to both TL-EN and STS dataset pairs. (3) then, we reduced the size of the model to offset the need for high computational resources. Based on our results, the models that were pre-trained to translation pairs and STS pairs can perform well for STSB task. Also, having it reduced to a smaller dimension has no negative effect on the performance but rather has a notable increase on the similarity scores. Moreover, models that were pre-trained to a similar dataset have a tremendous effect on the model’s performance scores.

Keywords: semantic matching, semantic textual similarity binary task, low resource minority language, fine-tuning, dimension reduction, transformer models

Procedia PDF Downloads 209

Authors: Sandy Black, Calum McLaughlin, Alessandro Pranzitelli, Marc Laing

Abstract:

Multiphase flow structures of two-phase multicomponent fluids were experimentally investigated in a large diameter high-pressure pipeline up to 130 bar at TÜV SÜD’s National Engineering Laboratory Advanced Multiphase Facility. One of the main objectives of the experimental test campaign was to evaluate the impact of pressure on multiphase flow patterns as much of the existing information is based on low-pressure measurements. The experiments were performed in a horizontal and vertical orientation in both 4-inch and 6-inch pipework using nitrogen, ExxsolTM D140 oil, and a 6% aqueous solution of NaCl at incremental pressures from 10 bar to 130 bar. To visualise the detailed structure of the flow of the entire cross-section of the pipe, a fast response X-ray tomography system was used. A wide range of superficial velocities from 0.6 m/s to 24.0 m/s for gas and 0.04 m/s and 6.48 m/s for liquid was examined to evaluate different flow regimes. The results illustrated the suppression of instabilities between the gas and the liquid at the measurement location and that intermittent or slug flow was observed less frequently as the pressure was increased. CFD modellings of low and high-pressure simulations were able to successfully predict the likelihood of intermittent flow; however, further tuning is necessary to predict the slugging frequency. The dataset generated is unique as limited datasets exist above 100 bar and is of considerable value to multiphase flow specialists and numerical modellers.

Keywords: computational fluid dynamics, high pressure, multiphase, X-ray tomography

Procedia PDF Downloads 142

Authors: Emily Awuor, Helga Huszar, Zoltan Bozoki

Abstract:

Isotope analysis has found numerous applications in the environmental science discipline, most common being the tracing of environmental contaminants on both regional and global scales. Many environmental contaminants contain ammonia (NH₃) since it is the most abundant gas in the atmosphere and its largest sources are from agricultural and industrial activities. NH₃ isotopes (¹⁴NH₃ and ¹⁵NH₃) are therefore important and can be used in the traceability studies of these atmospheric pollutants. The goal of the project is the construction of a photoacoustic spectroscopy system that is capable of measuring ¹⁵NH₃ isotope selectively in terms of its concentration. A further objective is for the system to be robust, easy-to-use, and automated. This is provided by using two telecommunication type near-infrared distributed feedback (DFB) diode lasers and a laser coupler as the light source in the photoacoustic measurement system. The central wavelength of the lasers in use was 1532 nm, with the tuning range of ± 1 nm. In this range, strong absorption lines can be found for both ¹⁴NH₃ and ¹⁵NH₃. For the selective measurement of ¹⁵NH₃, wavelengths were chosen where the cross effect of ¹⁴NH₃ and water vapor is negligible. We completed the calibration of the photoacoustic system, and as a result, the lowest detectable concentration was 3.32 ppm (3Ϭ) in the case of ¹⁵NH₃ and 0.44 ppm (3Ϭ) in the case of ¹⁴NH₃. The results are most useful in the environmental pollution measurement and analysis.

Keywords: ammonia isotope, near-infrared DFB diode laser, photoacoustic spectroscopy, environmental monitoring

Procedia PDF Downloads 144

Authors: Jessica Lopez Espejel, Mahaman Sanoussi Yahaya Alassan, Walid Dahhane, El Hassane Ettifouri

Abstract:

Pretrained transformer-based models have shown high performance in natural language generation tasks. However, a new wave of interest has surged: automatic programming language code generation. This task consists of translating natural language instructions to a source code. Despite the fact that well-known pre-trained models on language generation have achieved good performance in learning programming languages, effort is still needed in automatic code generation. In this paper, we introduce JaCoText, a model based on Transformer neural network. It aims to generate java source code from natural language text. JaCoText leverages the advantages of both natural language and code generation models. More specifically, we study some findings from state of the art and use them to (1) initialize our model from powerful pre-trained models, (2) explore additional pretraining on our java dataset, (3) lead experiments combining the unimodal and bimodal data in training, and (4) scale the input and output length during the fine-tuning of the model. Conducted experiments on CONCODE dataset show that JaCoText achieves new state-of-the-art results.

Keywords: java code generation, natural language processing, sequence-to-sequence models, transformer neural networks

Procedia PDF Downloads 283

Authors: Vinayambika S. Bhat, S. Shanmuga Priya, I. Thirunavukkarasu, Shreeranga Bhat

Abstract:

An attempt has been made to design a decoupling controller for systems with more inputs more outputs with dead time in it. The de-coupler is designed for the chemical process industry 3×3 plant transfer function with dead time. The Quantitative Feedback Theory (QFT) based controller has also been designed here for the 2×2 distillation column transfer function. The developed control techniques were simulated using the MATLAB/Simulink. Also, the stability of the process was analyzed, together with the presence of various perturbations in it. Time domain specifications like setting time along with overshoot and oscillations were analyzed to prove the efficiency of the de-coupler method. The load disturbance rejection was tested along with its performance. The QFT control technique was synthesized based on the stability and performance specifications in the presence of uncertainty in time constant of the plant transfer function through sequential loop shaping technique. Further, the energy efficiency of the distillation column was improved by proper tuning of the controller. A distillation column consumes 3% of the total energy consumption of the world. A suitable control technique is very important from an economic point of view. The real time implementation of the process is under process in our laboratory.

Keywords: distillation, energy, MIMO process, time delay, robust stability

Procedia PDF Downloads 413

Authors: Joshua Paul Steimel, Michael Pappas, Ethan Hall

Abstract:

Particle-particle interactions are critical in determining the state of an active matter system. Unique and ubiquitous non-equilibrium behavior like swarming, vortexing, spiraling, and much more is governed by interactions between active units or particles. In hybrid active-passive matter systems, the attraction between spinning active units in a 2D monolayer of passive particles is controlled by the mechanical behavior of the passive monolayer. We demonstrate here that the range and dynamics of this attraction can be controlled by changing the composition of the passive monolayer by adding dopant passive particles. These dopant passive particles effectively pin the movement of dislocation motion in the passive media and reduce the probability of defect motion required to erode the bridge of passive particles between active spinners, thus reducing the range of attraction. Additionally, by adding an out of plane component to the magnetic moment and creating a top-like motion a short range repulsion emerges between the top-like particle. At inter-top distances less than four particle diameters apart, the tops repel but beyond that, distance attract up to 13 particle diameters apart. The tops were also able to locally and transiently anneal the passive monolayer. Thus we demonstrate that by tuning several parameters of the hybrid active matter system, one can observe very different emergent behavior.

Keywords: active matter, colloids, ferromagnetic, annealing

Procedia PDF Downloads 108

Authors: Y. Alrubyli

Abstract:

Unmanned Arial Vehicles (UAVs), and more specifically, quadcopters need to be stable during their flights. Altitude stability is usually achieved by using a PID controller that is built into the flight controller software. Furthermore, the PID controller has gains that need to be tuned to reach optimal altitude stabilization during the quadcopter’s flight. For that, control system engineers need to tune those gains by using extensive modeling of the environment, which might change from one environment and condition to another. As quadcopters penetrate more sectors, from the military to the consumer sectors, they have been put into complex and challenging environments more than ever before. Hence, intelligent self-stabilizing quadcopters are needed to maneuver through those complex environments and situations. Here we show that by using online reinforcement learning with minimal background knowledge, the altitude stability of the quadcopter can be achieved using a model-free approach. We found that by using background knowledge instead of letting the online reinforcement learning algorithm wander for a while to tune the PID gains, altitude stabilization can be achieved faster. In addition, using this approach will accelerate development by avoiding extensive simulations before applying the PID gains to the real-world quadcopter. Our results demonstrate the possibility of using the trial and error approach of reinforcement learning combined with background knowledge to achieve faster quadcopter altitude stabilization in different environments and conditions.

Keywords: reinforcement learning, Q-leanring, online learning, PID tuning, unmanned aerial vehicle, quadcopter

Procedia PDF Downloads 170

Authors: Ricardo Mestre

Abstract:

The soundscape stands out as an agglomeration of sounds available in the world, associated with different contexts and origins, being a theme studied by various areas of knowledge, seeking to guide their benefits and their consequences, contributing to the welfare of society and other ecosystems. Murray Schafer, the author who originally developed this concept, highlights the need for a greater recognition of sound reality, through the selection and differentiation of sounds, contributing to a tuning of the world and to the balance and well-being of humanity. According to some authors sound environment, produced and created in various ways, provides various sources of information, contributing to the orientation of the human being, alerting and manipulating him during his daily journey, like small notifications received on a cell phone or other device with these features. In this way, it becomes possible to give sound its due importance in relation to the processes of individual representation, in manners of social, professional and emotional life. Ensuring an individual representation means providing the human being with new tools for the long process of reflection by recognizing his environment, the sounds that represent him, and his perspective on his respective function in it. In order to provide more information about the importance of the sound environment inherent to the individual reality, one introduces the term sound instance, in order to refer to the whole sound field existing in the individual's life, which is divided into four distinct subfields, but essential to the process of individual representation, called sound matrix, sound cycles, sound traces and sound interference.

Keywords: sound instance, soundscape, sound art, perception, composition

Procedia PDF Downloads 144

Authors: Maria C. Proença, Marco Aniceto, Pedro N. Santos, José C. Freitas

Abstract:

A methodology based on wavelets is proposed for the automatic location and delimitation of defects in limestone plates. Natural defects include dark colored spots, crystal zones trapped in the stone, areas of abnormal contrast colors, cracks or fracture lines, and fossil patterns. Although some of these may or may not be considered as defects according to the intended use of the plate, the goal is to pair each stone with a map of defects that can be overlaid on a computer display. These layers of defects constitute a database that will allow the preliminary selection of matching tiles of a particular variety, with specific dimensions, for a requirement of N square meters, to be done on a desktop computer rather than by a two-hour search in the storage park, with human operators manipulating stone plates as large as 3 m x 2 m, weighing about one ton. Accident risks and work times are reduced, with a consequent increase in productivity. The base for the algorithm is wavelet decomposition executed in two instances of the original image, to detect both hypotheses – dark and clear defects. The existence and/or size of these defects are the gauge to classify the quality grade of the stone products. The tuning of parameters that are possible in the framework of the wavelets corresponds to different levels of accuracy in the drawing of the contours and selection of the defects size, which allows for the use of the map of defects to cut a selected stone into tiles with minimum waste, according the dimension of defects allowed.

Keywords: automatic detection, defects, fracture lines, wavelets

Procedia PDF Downloads 246

Authors: Simeon Mayala, Ida Herdlevær, Jonas Bull Haugsøen, Shamundeeswari Anandan, Sonia Gavasso, Morten Brun

Abstract:

In this paper, we propose a minimum spanning tree-based method for segmenting brain tumors. The proposed method performs interactive segmentation based on the minimum spanning tree without tuning parameters. The steps involve preprocessing, making a graph, constructing a minimum spanning tree, and a newly implemented way of interactively segmenting the region of interest. In the preprocessing step, a Gaussian filter is applied to 2D images to remove the noise. Then, the pixel neighbor graph is weighted by intensity differences and the corresponding minimum spanning tree is constructed. The image is loaded in an interactive window for segmenting the tumor. The region of interest and the background are selected by clicking to split the minimum spanning tree into two trees. One of these trees represents the region of interest and the other represents the background. Finally, the segmentation given by the two trees is visualized. The proposed method was tested by segmenting two different 2D brain T1-weighted magnetic resonance image data sets. The comparison between our results and the standard gold segmentation confirmed the validity of the minimum spanning tree approach. The proposed method is simple to implement and the results indicate that it is accurate and efficient.

Keywords: brain tumor, brain tumor segmentation, minimum spanning tree, segmentation, image processing

Procedia PDF Downloads 119

Authors: Karla Lienau, Rafael Müller, René Moré, Debora Ressnig, Dan Cook, Richard Walton, Greta R. Patzke

Abstract:

The increasing demand for renewable energy sources and storable fuels underscores the high potential of artificial photosynthesis. The four electron transfer process of water oxidation remains the bottleneck of water splitting, so that special emphasis is placed on the development of economic, stable and efficient water oxidation catalysts (WOCs). Our investigations introduced cobalt carbodiimide CoNCN and its transition metal analogues as WOC types, and further studies are focused on the interaction of different transition metals in the convenient all-nitrogen/carbon matrix. This provides further insights into the nature of the ‘true catalyst’ for cobalt centers in this non-oxide environment. Water oxidation activity is evaluated with complementary methods, namely photocatalytically using a Ru-dye sensitized standard setup as well as electrocatalytically, via immobilization of the WOCs on glassy carbon electrodes. To further explore the tuning potential of transition metal combinations, complementary investigations were carried out in oxidic spinel WOC matrices with more versatile host options than the carbodiimide framework. The influence of the preparative history on the WOC performance was evaluated with different synthetic methods (e.g. hydrothermally or microwave assisted). Moreover, the growth mechanism of nanoscale Co3O4-spinel as a benchmark WOC was investigated with in-situ PXRD techniques.

Keywords: carbodiimide, photocatalysis, spinels, water oxidation

Procedia PDF Downloads 288

Authors: Eiman Kattan

Abstract:

This study aims to evaluate the impact of classifying labelled remote sensing images conventional neural network (CNN) architecture, i.e., AlexNet on different land cover scenarios based on two remotely sensed datasets from different point of views such as the computational time and performance. Thus, a set of experiments were conducted to specify the effectiveness of the selected convolutional neural network using two implementing approaches, named fully trained and fine-tuned. For validation purposes, two remote sensing datasets, AID, and RSSCN7 which are publicly available and have different land covers features were used in the experiments. These datasets have a wide diversity of input data, number of classes, amount of labelled data, and texture patterns. A specifically designed interactive deep learning GPU training platform for image classification (Nvidia Digit) was employed in the experiments. It has shown efficiency in training, validation, and testing. As a result, the fully trained approach has achieved a trivial result for both of the two data sets, AID and RSSCN7 by 73.346% and 71.857% within 24 min, 1 sec and 8 min, 3 sec respectively. However, dramatic improvement of the classification performance using the fine-tuning approach has been recorded by 92.5% and 91% respectively within 24min, 44 secs and 8 min 41 sec respectively. The represented conclusion opens the opportunities for a better classification performance in various applications such as agriculture and crops remote sensing.

Keywords: conventional neural network, remote sensing, land cover, land use

Procedia PDF Downloads 369

Authors: Xiaolin Zeng, Lei Cheng, Zhirong Li, Xueping Liu

Abstract:

Aiming at the problem of low accuracy and susceptibility to cultural relic diseases in the generation of high-resolution archaeology drawings by current image generation algorithms, an archaeology drawings generation algorithm based on a conditional generative adversarial network is proposed. An attention mechanism is added into the high-resolution image generation network as the backbone network, which enhances the line feature extraction capability and improves the accuracy of line drawing generation. A dual-branch parallel architecture consisting of two backbone networks is implemented, where the semantic translation branch extracts semantic features from orthophotographs of cultural relics, and the gradient screening branch extracts effective gradient features. Finally, the fusion fine-tuning module combines these two types of features to achieve the generation of high-quality and high-resolution archaeology drawings. Experimental results on the self-constructed archaeology drawings dataset of grotto temple statues show that the proposed algorithm outperforms current mainstream image generation algorithms in terms of pixel accuracy (PA), structural similarity (SSIM), and peak signal-to-noise ratio (PSNR) and can be used to assist in drawing archaeology drawings.

Keywords: archaeology drawings, digital heritage, image generation, deep learning

Procedia PDF Downloads 56

Authors: P. S. Jagadeesh Kumar, Yang Yung, Wenli Hu

Abstract:

Speech emotion classification is a dominant research field in finding a sturdy and profligate classifier appropriate for different real-life applications. This effort accentuates on classifying different emotions from speech signal quarried from the features related to pitch, formants, energy contours, jitter, shimmer, spectral, perceptual and temporal features. Tensor deep stacking neural networks were supported to examine the factors that influence the classification success rate. Facial electromyography signals were composed of several forms of focuses in a controlled atmosphere by means of audio-visual stimuli. Proficient facial electromyography signals were pre-processed using moving average filter, and a set of arithmetical features were excavated. Extracted features were mapped into consistent emotions using bilinear mapping. With facial electromyography signals, a database comprising diverse emotions will be exposed with a suitable fine-tuning of features and training data. A success rate of 92% can be attained deprived of increasing the system connivance and the computation time for sorting diverse emotional states.

Keywords: speech emotion classification, tensor deep stacking neural networks, facial electromyography, bilinear mapping, audio-visual stimuli

Procedia PDF Downloads 253

Authors: Qisheng Hu, Jianglei Han, Yue Yang, My Hoa Ha

Abstract:

When a company is considering purchasing software for commercial use, contract risk assessment is critical to identify risks to mitigate the potential adverse business impact, e.g., security, financial and regulatory risks. Contract risk assessment requires reviewers with specialized knowledge and time to evaluate the legal documents manually. Specifically, validating contracts for a software vendor requires the following steps: manual screening, interpreting legal documents, and extracting risk-prone segments. To automate the process, we proposed a framework to assist legal contract document risk identification, leveraging pre-trained deep learning models and natural language processing techniques. Given a set of pre-defined risk evaluation problems, our framework utilizes the pre-trained transformer-based models for question-answering to identify risk-prone sections in a contract. Furthermore, the question-answering model encodes the concatenated question-contract text and predicts the start and end position for clause extraction. Due to the limited labelled dataset for training, we leveraged transfer learning by fine-tuning the models with the CUAD dataset to enhance the model. On a dataset comprising 287 contract documents and 2000 labelled samples, our best model achieved an F1 score of 0.687.

Keywords: contract risk assessment, NLP, transfer learning, question answering

Procedia PDF Downloads 128

Authors: Ahmed Al-Mansour, Qiang Zeng

Abstract:

Recycling waste plastics in the form of concrete components, i.e. fine aggregates, has been an attractive topic among the society of civil engineers. Not only does the recycling of plastics reduce the overall cost of concrete production, but it also takes part in solving environmental issues. Nevertheless, the incorporation of recycled plastics into concrete results in an increasing reduction in the mechanical properties of concrete as the percentage of replacement of natural aggregates increases. In order to overcome this reduction, Ethylene-vinyl acetate (EVA) was used as an additive in concrete with recycled plastic aggregates. The aim of this additive is to: 1) increase the interfacial interaction at the interfacial transition zone (ITZ) between plastic pellets and cement matrix, and 2) mitigate the loss in mechanical properties. Three different groups of samples (i.e. cubes and prisms) were tested according to the plastics substituting fine aggregates. 5, 10, and 15% of fine aggregates were substituted for recycled plastic pellets, and 2 – 4% of the cement was substituted for EVA that produces a flexible agent when mixed properly with water. Compressive and tensile strength tests were conducted for the mechanical properties, while SEM and X-CT scan were implemented for further investigation of calcium-silicate-hydrate (C–S–H) formation and ITZ analysis. The optimal amount of plastic particles with EVA is suggested to get the most compact and dense matrix structure according to the results of this study.

Keywords: the durability of concrete, ethylene-vinyl acetate (EVA), interfacial transition zone (ITZ), recycled plastics

Procedia PDF Downloads 184

Authors: Ken Hampshire, Thomas Mazzuchi, Shahram Sarkani

Abstract:

As a subset of metaheuristics, nature-inspired optimization algorithms such as particle swarm optimization (PSO) have shown promise both in solving intractable problems and in their extensibility to novel problem formulations due to their general approach requiring few assumptions. Unfortunately, single instantiations of algorithms require detailed tuning of parameters and cannot be proven to be best suited to a particular illustrative problem on account of the “no free lunch” (NFL) theorem. Using these algorithms in real-world problems requires exquisite knowledge of the many techniques and is not conducive to reconciling the various approaches to given classes of problems. This research aims to present a unified view of PSO-based approaches from the perspective of relevant systems engineering problems, with the express purpose of then eliciting the best solution for any problem formulation in an ensemble learning bucket of models approach. The central hypothesis of the research is that extending the PSO algorithms found in the literature to real-world optimization problems requires a general ensemble-based method for all problem formulations but a specific implementation and solution for any instance. The main results are a problem-based literature survey and a general method to find more globally optimal solutions for any systems engineering optimization problem.

Keywords: particle swarm optimization, nature-inspired optimization, metaheuristics, systems engineering, ensemble learning

Procedia PDF Downloads 96

Authors: Ronal Muresano, Andrea Pagano

Abstract:

Nowadays, the mathematical/statistical applications are developed with more complexity and accuracy. However, these precisions and complexities have brought as result that applications need more computational power in order to be executed faster. In this sense, the multicore environments are playing an important role to improve and to optimize the execution time of these applications. These environments allow us the inclusion of more parallelism inside the node. However, to take advantage of this parallelism is not an easy task, because we have to deal with some problems such as: cores communications, data locality, memory sizes (cache and RAM), synchronizations, data dependencies on the model, etc. These issues are becoming more important when we wish to improve the application’s performance and scalability. Hence, this paper describes an optimization method developed for Systemic Model of Banking Originated Losses (SYMBOL) tool developed by the European Commission, which is based on analyzing the application's weakness in order to exploit the advantages of the multicore. All these improvements are done in an automatic and transparent manner with the aim of improving the performance metrics of our tool. Finally, experimental evaluations show the effectiveness of our new optimized version, in which we have achieved a considerable improvement on the execution time. The time has been reduced around 96% for the best case tested, between the original serial version and the automatic parallel version.

Keywords: algorithm optimization, bank failures, OpenMP, parallel techniques, statistical tool

Procedia PDF Downloads 366

Authors: Vladimir Mantsevich, Natalya Maslova, Petr Arseyev

Abstract:

We investigated the tunneling current peculiarities in the system of two coupled by means of the external field quantum dots (QDs) weakly connected to the electrodes in the presence of Coulomb correlations between localized electrons by means of Heisenberg equations for pseudo operators with constraint. Special role of multi-electronic states was demonstrated. Various single-electron levels location relative to the sample Fermi level and to the applied bias value in symmetric tunneling contact were investigated. Rabi frequency tuning results in the single-electron energy levels spacing. We revealed the appearance of negative tunneling conductivity and demonstrated multiple switching "on" and "off" of the tunneling current depending on the Coulomb correlations value, Rabi frequency amplitude and energy levels spacing. We proved that Coulomb correlations strongly influence the system behavior. We demonstrated the presence of multi-stability in the coupled QDs with Coulomb correlations when single value of the tunneling current amplitude corresponds to the two values of Rabi frequency in the case when both single-electron energy levels are located slightly above eV and are close to each other. This effect disappears when the single-electron energy levels spacing increases.

Keywords: Coulomb correlations, negative tunneling conductivity, quantum dots, rabi frequency

Procedia PDF Downloads 450

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 12

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 74

Authors: D. Baratian, A. Cavalli, D. van den Ende, F. Mugele

Abstract:

The morphology of liquid bridges inside complex geometries is the subject of interest for many years. These efforts try to find stable liquid configuration considering the boundary condition and the physical properties of the system. On the other hand precise manipulation of droplets is highly significant in many microfluidic applications. The liquid configuration in a complex geometry can be switched by means of external stimuli. We show manipulation of droplets in a wedge structure. The profile and position of a drop in a wedge geometry has been calculated analytically assuming negligible contact angle hysteresis. The characteristic length of liquid bridge and its interfacial tension inside the surrounding medium along with the geometrical parameters of the system determine the morphology and equilibrium position of drop in the system. We use electrowetting to modify one the governing parameters to manipulate the droplet. Electrowetting provides the capability to have precise control on the drop position through tuning the voltage and consequently changing the contact angle. This technique is employed to tune drop displacement and control its position inside the wedge. Experiments demonstrate precise drop movement to its predefined position inside the wedge geometry. Experimental results show promising consistency as it is compared to our geometrical model predictions. For such a drop manipulation, appealing applications in microfluidics have been considered.

Keywords: liquid bridges, microfluidics, drop manipulation, wetting, electrowetting, capillarity

Procedia PDF Downloads 475

Authors: Dongju Kim, Youngjoo Suh, Hyojin Kim, Gyeongyeong Kim

Abstract:

Acoustic Emission Testing (AET) is widely used to test the structural integrity of an operational hydrogen storage container, and clustering algorithms are frequently used in pattern recognition methods to interpret AET results. However, the interpretation of AET results can vary from user to user as the tuning of the relevant parameters relies on the user's experience and knowledge of AET. Therefore, it is necessary to use a deep learning model to identify patterns in acoustic emission (AE) signal data that can be used to classify defects instead. In this paper, a deep learning-based model for classifying the types of defects in hydrogen storage tanks, using AE sensor waveforms, is proposed. As hydrogen storage tanks are commonly constructed using carbon fiber reinforced polymer composite (CFRP), a defect classification dataset is collected through a tensile test on a specimen of CFRP with an AE sensor attached. The performance of the classification model, using one-dimensional convolutional neural network (1-D CNN) and synthetic minority oversampling technique (SMOTE) data augmentation, achieved 91.09% accuracy for each defect. It is expected that the deep learning classification model in this paper, used with AET, will help in evaluating the operational safety of hydrogen storage containers.

Keywords: acoustic emission testing, carbon fiber reinforced polymer composite, one-dimensional convolutional neural network, smote data augmentation

Procedia PDF Downloads 93

Authors: Chiraz Ben Jabeur, Hassene Seddik

Abstract:

In this paper, a full approach of modeling and control of a four-rotor unmanned air vehicle (UAV), known as quad-rotor aircraft, is presented. In fact, a PD and a PD optimized Neural Networks Approaches (PD-NN) are developed to be applied to control a quad-rotor. The goal of this work is to concept a smart self-tuning PD controller based on neural networks able to supervise the quad-rotor for an optimized behavior while tracking the desired trajectory. Many challenges could arise if the quad-rotor is navigating in hostile environments presenting irregular disturbances in the form of wind added to the model on each axis. Thus, the quad-rotor is subject to three-dimensional unknown static/varying wind disturbances. The quad-rotor has to quickly perform tasks while ensuring stability and accuracy and must behave rapidly with regard to decision-making facing disturbances. This technique offers some advantages over conventional control methods such as PD controller. Simulation results are obtained with the use of Matlab/Simulink environment and are founded on a comparative study between PD and PD-NN controllers based on wind disturbances. These later are applied with several degrees of strength to test the quad-rotor behavior. These simulation results are satisfactory and have demonstrated the effectiveness of the proposed PD-NN approach. In fact, this controller has relatively smaller errors than the PD controller and has a better capability to reject disturbances. In addition, it has proven to be highly robust and efficient, facing turbulences in the form of wind disturbances.

Keywords: hostile environment, PD and PD-NN controllers, quad-rotor control, robustness against disturbance

Procedia PDF Downloads 136

Authors: Larry Powell, Seth Polsley, Drew Casey, Tracy Hammond

Abstract:

Human activity recognition (HAR) systems have shown positive performance when recognizing repetitive activities like walking, running, and sleeping. Water-based activities are a reasonably new area for activity recognition. However, water-based activity recognition has largely focused on supporting the elite and competitive swimming population, which already has amazing coordination and proper form. Beginner swimmers are not perfect, and activity recognition needs to support the individual motions to help beginners. Activity recognition algorithms are traditionally built around short segments of timed sensor data. Using a time window input can cause performance issues in the machine learning model. The window’s size can be too small or large, requiring careful tuning and precise data segmentation. In this work, we present a method that uses a time window as the initial segmentation, then separates the data based on the change in the sensor value. Our system uses a multi-phase segmentation method that pulls all peaks and valleys for each axis of an accelerometer placed on the swimmer’s lower back. This results in high recognition performance using leave-one-subject-out validation on our study with 20 beginner swimmers, with our model optimized from our final dataset resulting in an F-Score of 0.95.

Keywords: time window, peak/valley segmentation, feature extraction, beginner swimming, activity recognition

Procedia PDF Downloads 121

Authors: Isaac Laryea

Abstract:

Underwater remote-operated vehicles (ROVs) are highly useful in aquatic research and underwater operations. Unfortunately, unsteady and unpredictable conditions underwater make it difficult for underwater vehicles to maintain a steady attitude during motion. Existing underwater vehicles make use of multiple thrusters positioned at specific positions on their frame to maintain a certain pose. This study proposes an alternate way of maintaining a steady attitude during horizontal motion at low speeds by making use of a thrust vector-controlled propulsion system. The study began by carrying out some preliminary calculations to get an idea of a suitable shape and form factor. Flow simulations were carried out to ensure that enough thrust could be generated to move the system. Using the Lagrangian approach, a mathematical system was developed for the ROV, and this model was used to design a control system. A PID controller was selected for the control system. However, after tuning, it was realized that a PD controller satisfied the design specifications. The designed control system produced an overshoot of 6.72%, with a settling time of 0.192s. To achieve the effect of thrust vectoring, an inverse kinematics synthesis was carried out to determine what angle the actuators need to move to. After building the system, intermittent angular displacements of 10°, 15°, and 20° were given during bench testing, and the response of the control system as well as the servo motor angle was plotted. The final design was able to move in water but was not able to handle large angular displacements as a result of the small angle approximation used in the mathematical model.

Keywords: PID control, thrust vectoring, parallel manipulators, ROV, underwater, attitude control

Procedia PDF Downloads 66

Authors: Mohammad Amin Tutunchian, Pedram Roshani, Reza Rezvani, Julio Ángel Infante Sedano

Abstract:

The highly recommended approach to design, known as the load and resistance factor design (LRFD) method, employs the geotechnical resistance factor (GRF) for shaping pile foundation designs. Within the standard process for designing pile foundations, geotechnical engineers commonly adopt a design strategy rooted in saturated soil mechanics (SSM), often disregarding the impact of unsaturated soil behavior. This oversight within the design procedure leads to the omission of the enhancement in shear strength exhibited by unsaturated soils, resulting in a more cautious outcome in design results. This research endeavors to present a methodology for fine-tuning the GRF used for axially loaded driven piles in Winnipeg, Canada. This is achieved through the application of a well-established probabilistic approach known as the first-order second moment (FOSM) method while also accounting for the influence of unsaturated soil behavior. The findings of this study demonstrate that incorporating the influence of unsaturated conditions yields an elevation in projected bearing capacity and recommends higher GRF values in accordance with established codes. Additionally, a novel factor referred to as phy has been introduced to encompass the impact of saturation conditions in the calculation of pile bearing capacity, as guided by prevalent static analysis techniques.

Keywords: unsaturated soils, shear strength, LRFD, FOSM, GRF

Procedia PDF Downloads 88

Authors: Kristian Bautista, Ruben A. Idoy

Abstract:

A bio-inspired metaheuristic algorithm inspired by the theory of collective animal behavior (CAB) was integrated to density-based clustering modeled as multimodal optimization problem. The algorithm was tested on synthetic, Iris, Glass, Pima and Thyroid data sets in order to measure its effectiveness relative to CDE-based Clustering algorithm. Upon preliminary testing, it was found out that one of the parameter settings used was ineffective in performing clustering when applied to the algorithm prompting the researcher to do an investigation. It was revealed that fine tuning distance δ3 that determines the extent to which a given data point will be clustered helped improve the quality of cluster output. Even though the modification of distance δ3 significantly improved the solution quality and cluster output of the algorithm, results suggest that there is no difference between the population mean of the solutions obtained using the original and modified parameter setting for all data sets. This implies that using either the original or modified parameter setting will not have any effect towards obtaining the best global and local animal positions. Results also suggest that CDE-based clustering algorithm is better than CAB-density clustering algorithm for all data sets. Nevertheless, CAB-density clustering algorithm is still a good clustering algorithm because it has correctly identified the number of classes of some data sets more frequently in a thirty trial run with a much smaller standard deviation, a potential in clustering high dimensional data sets. Thus, the researcher recommends further investigation in the post-processing stage of the algorithm.

Keywords: clustering, metaheuristics, collective animal behavior algorithm, density-based clustering, multimodal optimization

Procedia PDF Downloads 230

Authors: Douglas Yeboah, Jai Singh

Abstract:

It has been experimentally demonstrated that exciton diffusion length in organic solids can be improved by fine-tuning the material parameters that govern exciton transfer. Here, a theoretical study is carried out to support this finding. We have therefore derived expressions for the exciton diffusion length and diffusion coefficient of singlet and triplet excitons using Förster resonance energy transfer and Dexter carrier transfer mechanisms and are plotted as a function of photoluminescence (PL) quantum yield, spectral overlap integral, refractive index and dipole moment of the photoactive material. We found that singlet exciton diffusion length increases with PL quantum yield and spectral overlap integral, and decreases with increase in refractive index. Likewise, the triplet exciton diffusion length increases when PL quantum yield increases and dipole moment decreases. The calculated diffusion lengths in different organic materials are compared with existing experimental values and found to be in reasonable agreement. The results are expected to provide insight in developing new organic materials for fabricating bulk heterojunction (BHJ) organic solar cells (OSCs) with better photoconversion efficiency.

Keywords: Dexter carrier transfer, diffusion coefficient, exciton diffusion length, Föster resonance energy transfer, photoactive materials, photophysical parameters

Procedia PDF Downloads 331

Authors: Ankit Sinha, Soham Banerjee, Pratik Chattopadhyay

Abstract:

Automated product recognition in retail stores is an important real-world application in the domain of Computer Vision and Pattern Recognition. In this paper, we consider the problem of automatically identifying the classes of the products placed on racks in retail stores from an image of the rack and information about the query/product images. We improve upon the existing approaches in terms of effectiveness and memory requirement by developing a two-stage object detection and recognition pipeline comprising of a Faster-RCNN-based object localizer that detects the object regions in the rack image and a ResNet-18-based image encoder that classifies the detected regions into the appropriate classes. Each of the models is fine-tuned using appropriate data sets for better prediction and data augmentation is performed on each query image to prepare an extensive gallery set for fine-tuning the ResNet-18-based product recognition model. This encoder is trained using a triplet loss function following the strategy of online-hard-negative-mining for improved prediction. The proposed models are lightweight and can be connected in an end-to-end manner during deployment to automatically identify each product object placed in a rack image. Extensive experiments using Grozi-32k and GP-180 data sets verify the effectiveness of the proposed model.

Keywords: retail stores, faster-RCNN, object localization, ResNet-18, triplet loss, data augmentation, product recognition

Procedia PDF Downloads 154