Search results for: representation of graph models
Commenced in January 2007
Frequency: Monthly
Edition: International
Paper Count: 8183

Search results for: representation of graph models

7313 Investigating the performance of machine learning models on PM2.5 forecasts: A case study in the city of Thessaloniki

Authors: Alexandros Pournaras, Anastasia Papadopoulou, Serafim Kontos, Anastasios Karakostas

Abstract:

The air quality of modern cities is an important concern, as poor air quality contributes to human health and environmental issues. Reliable air quality forecasting has, thus, gained scientific and governmental attention as an essential tool that enables authorities to take proactive measures for public safety. In this study, the potential of Machine Learning (ML) models to forecast PM2.5 at local scale is investigated in the city of Thessaloniki, the second largest city in Greece, which has been struggling with the persistent issue of air pollution. ML models, with proven ability to address timeseries forecasting, are employed to predict the PM2.5 concentrations and the respective Air Quality Index 5-days ahead by learning from daily historical air quality and meteorological data from 2014 to 2016 and gathered from two stations with different land use characteristics in the urban fabric of Thessaloniki. The performance of the ML models on PM2.5 concentrations is evaluated with common statistical methods, such as R squared (r²) and Root Mean Squared Error (RMSE), utilizing a portion of the stations’ measurements as test set. A multi-categorical evaluation is utilized for the assessment of their performance on respective AQIs. Several conclusions were made from the experiments conducted. Experimenting on MLs’ configuration revealed a moderate effect of various parameters and training schemas on the model’s predictions. Their performance of all these models were found to produce satisfactory results on PM2.5 concentrations. In addition, their application on untrained stations showed that these models can perform well, indicating a generalized behavior. Moreover, their performance on AQI was even better, showing that the MLs can be used as predictors for AQI, which is the direct information provided to the general public.

Keywords: Air Quality, AQ Forecasting, AQI, Machine Learning, PM2.5

Procedia PDF Downloads 79
7312 Quantitative Structure-Activity Relationship Study of Some Quinoline Derivatives as Antimalarial Agents

Authors: M. Ouassaf, S. Belaid

Abstract:

A series of quinoline derivatives with antimalarial activity were subjected to two-dimensional quantitative structure-activity relationship (2D-QSAR) studies. Three models were implemented using multiple regression linear MLR, a regression partial least squares (PLS), nonlinear regression (MNLR), to see which descriptors are closely related to the activity biologic. We relied on a principal component analysis (PCA). Based on our results, a comparison of the quality of, MLR, PLS, and MNLR models shows that the MNLR (R = 0.914 and R² = 0.835, RCV= 0.853) models have substantially better predictive capability because the MNLR approach gives better results than MLR (R = 0.835 and R² = 0,752, RCV=0.601)), PLS (R = 0.742 and R² = 0.552, RCV=0.550) The model of MNLR gave statistically significant results and showed good stability to data variation in leave-one-out cross-validation. The obtained results suggested that our proposed model MNLR may be useful to predict the biological activity of derivatives of quinoline.

Keywords: antimalarial, quinoline, QSAR, PCA, MLR , MNLR, MLR

Procedia PDF Downloads 157
7311 An Adaptive Hybrid Surrogate-Assisted Particle Swarm Optimization Algorithm for Expensive Structural Optimization

Authors: Xiongxiong You, Zhanwen Niu

Abstract:

Choosing an appropriate surrogate model plays an important role in surrogates-assisted evolutionary algorithms (SAEAs) since there are many types and different kernel functions in the surrogate model. In this paper, an adaptive selection of the best suitable surrogate model method is proposed to solve different kinds of expensive optimization problems. Firstly, according to the prediction residual error sum of square (PRESS) and different model selection strategies, the excellent individual surrogate models are integrated into multiple ensemble models in each generation. Then, based on the minimum root of mean square error (RMSE), the best suitable surrogate model is selected dynamically. Secondly, two methods with dynamic number of models and selection strategies are designed, which are used to show the influence of the number of individual models and selection strategy. Finally, some compared studies are made to deal with several commonly used benchmark problems, as well as a rotor system optimization problem. The results demonstrate the accuracy and robustness of the proposed method.

Keywords: adaptive selection, expensive optimization, rotor system, surrogates assisted evolutionary algorithms

Procedia PDF Downloads 141
7310 Material Response Characterisation of a PolyJet 3D Printed Human Infant Skull

Authors: G. A. Khalid, R. Prabhu, W. Whittington, M. D. Jones

Abstract:

To establish a causal relationship of infant head injury consequences, this present study addresses the necessary challenges of cranial geometry and the physical response complexities of the paediatric head tissues. Herein, we describe a new approach to characterising and understanding infant head impact mechanics by developing printed head models, using high resolution clinical postmortem imaging, to provide the most complete anatomical representation currently available, and biological material response data-matched polypropylene polymers, to replicate the relative mechanical response properties of immature cranial bone, sutures and fontanelles. Additive manufacturing technology was applied to creating a physical polymeric model of a newborn infant skull, using PolyJet printed materials. Infant skull materials responses, were matched by a response characterisation study, utilising uniaxial tensile testing (1 mm min-1 loading rate), to determine: the stiffness, ultimate tensile strength and maximum strain of rigid and rubber additively manufactured acrylates. The results from the mechanical experiments confirm that the polymeric materials RGD835 Vero White Plus (White), representing the frontal and parietal bones; RGD8510- DM Rigid Light Grey25 (Grey), representing the occipital bone; and FLX9870-DM (Black) representing the suture and fontanelles, were found to show a close stiffness -correlation (E) at ambient temperatures. A 3D physical model of infant head was subsequently printed from the matched materials and subsequently validated against results obtained from a series of Post Mortem Human Surrogate (PMHS) tests. A close correlation was demonstrated between the model impact tests and the PMHS. This study, therefore, represents a key step towards applying printed physical models to understanding head injury biomechanics and is useful in the efforts to predict and mitigate head injury consequences in infants, whether accidental or by abuse.

Keywords: infant head trauma, infant skull, material response, post mortem human subjects, polyJet printing

Procedia PDF Downloads 141
7309 Importance of Solubility and Bubble Pressure Models to Predict Pressure of Nitrified Oil Based Drilling Fluid in Dual Gradient Drilling

Authors: Sajjad Negahban, Ruihe Wang, Baojiang Sun

Abstract:

Gas-lift dual gradient drilling is a solution for deepwater drilling challenges. As well, Continuous development of drilling technology leads to increase employment of mineral oil based drilling fluids and synthetic-based drilling fluids, which have adequate characteristics such as: high rate of penetration, lubricity, shale inhibition and low toxicity. The paper discusses utilization of nitrified mineral oil base drilling for deepwater drilling and for more accurate prediction of pressure in DGD at marine riser, solubility and bubble pressure were considered in steady state hydraulic model. The Standing bubble pressure and solubility correlations, and two models which were acquired from experimental determination were applied in hydraulic model. The effect of the black oil correlations, and new solubility and bubble pressure models was evaluated on the PVT parameters such as oil formation volume factor, density, viscosity, volumetric flow rate. Eventually, the consequent simulated pressure profile due to these models was presented.

Keywords: solubility, bubble pressure, gas-lift dual gradient drilling, steady state hydraulic model

Procedia PDF Downloads 278
7308 Personal Information Classification Based on Deep Learning in Automatic Form Filling System

Authors: Shunzuo Wu, Xudong Luo, Yuanxiu Liao

Abstract:

Recently, the rapid development of deep learning makes artificial intelligence (AI) penetrate into many fields, replacing manual work there. In particular, AI systems also become a research focus in the field of automatic office. To meet real needs in automatic officiating, in this paper we develop an automatic form filling system. Specifically, it uses two classical neural network models and several word embedding models to classify various relevant information elicited from the Internet. When training the neural network models, we use less noisy and balanced data for training. We conduct a series of experiments to test my systems and the results show that our system can achieve better classification results.

Keywords: artificial intelligence and office, NLP, deep learning, text classification

Procedia PDF Downloads 202
7307 Marriage Domination and Divorce Domination in Graphs

Authors: Mark L. Caay, Rodolfo E. Maza

Abstract:

In this paper, the authors define two new variants of domination in graphs: the marriage and the divorce domination. A subset S ⊆ V (G) is said to be a marriage dominating set of G if for every e ∈ E(G), there exists a u ∈ V (G) such that u is one of the end vertex of e. A marriage dominating set S ⊆ V (G) is said to be a divorce dominating set of G if G\S is a disconnected graph. In this study, the authors present conditions of graphs for which the marriage and the divorce domination will take place and for which the two sets will coincide. Furthermore, the author gives the necessary and sufficient conditions for marriage domination to avoid divorce.

Keywords: domination, decomposition, marriage domination, divorce domination, marriage theorem

Procedia PDF Downloads 25
7306 Validation and Projections for Solar Radiation up to 2100: HadGEM2-AO Global Circulation Model

Authors: Elison Eduardo Jardim Bierhals, Claudineia Brazil, Deivid Pires, Rafael Haag, Elton Gimenez Rossini

Abstract:

The objective of this work is to evaluate the results of solar radiation projections between 2006 and 2013 for the state of Rio Grande do Sul, Brazil. The projections are provided by the General Circulation Models (MCGs) belonging to the Coupled Model Intercomparison Phase 5 (CMIP5). In all, the results of the simulation of six models are evaluated, compared to monthly data, measured by a network of thirteen meteorological stations of the National Meteorological Institute (INMET). The performance of the models is evaluated by the Nash coefficient and the Bias. The results are presented in the form of tables, graphs and spatialization maps. The ACCESS1-0 RCP 4.5 model presented the best results for the solar radiation simulations, for the most optimistic scenario, in much of the state. The efficiency coefficients (CEF) were between 0.95 and 0.98. In the most pessimistic scenario, HADGen2-AO RCP 8.5 had the best accuracy among the analyzed models, presenting coefficients of efficiency between 0.94 and 0.98. From this validation, solar radiation projection maps were elaborated, indicating a seasonal increase of this climatic variable in some regions of the Brazilian territory, mainly in the spring.

Keywords: climate change, projections, solar radiation, validation

Procedia PDF Downloads 209
7305 A Neural Network Classifier for Estimation of the Degree of Infestation by Late Blight on Tomato Leaves

Authors: Gizelle K. Vianna, Gabriel V. Cunha, Gustavo S. Oliveira

Abstract:

Foliage diseases in plants can cause a reduction in both quality and quantity of agricultural production. Intelligent detection of plant diseases is an essential research topic as it may help monitoring large fields of crops by automatically detecting the symptoms of foliage diseases. This work investigates ways to recognize the late blight disease from the analysis of tomato digital images, collected directly from the field. A pair of multilayer perceptron neural network analyzes the digital images, using data from both RGB and HSL color models, and classifies each image pixel. One neural network is responsible for the identification of healthy regions of the tomato leaf, while the other identifies the injured regions. The outputs of both networks are combined to generate the final classification of each pixel from the image and the pixel classes are used to repaint the original tomato images by using a color representation that highlights the injuries on the plant. The new images will have only green, red or black pixels, if they came from healthy or injured portions of the leaf, or from the background of the image, respectively. The system presented an accuracy of 97% in detection and estimation of the level of damage on the tomato leaves caused by late blight.

Keywords: artificial neural networks, digital image processing, pattern recognition, phytosanitary

Procedia PDF Downloads 330
7304 Stock Price Prediction Using Time Series Algorithms

Authors: Sumit Sen, Sohan Khedekar, Umang Shinde, Shivam Bhargava

Abstract:

This study has been undertaken to investigate whether the deep learning models are able to predict the future stock prices by training the model with the historical stock price data. Since this work required time series analysis, various models are present today to perform time series analysis such as Recurrent Neural Network LSTM, ARIMA and Facebook Prophet. Applying these models the movement of stock price of stocks are predicted and also tried to provide the future prediction of the stock price of a stock. Final product will be a stock price prediction web application that is developed for providing the user the ease of analysis of the stocks and will also provide the predicted stock price for the next seven days.

Keywords: Autoregressive Integrated Moving Average, Deep Learning, Long Short Term Memory, Time-series

Procedia PDF Downloads 143
7303 In-Context Meta Learning for Automatic Designing Pretext Tasks for Self-Supervised Image Analysis

Authors: Toktam Khatibi

Abstract:

Self-supervised learning (SSL) includes machine learning models that are trained on one aspect and/or one part of the input to learn other aspects and/or part of it. SSL models are divided into two different categories, including pre-text task-based models and contrastive learning ones. Pre-text tasks are some auxiliary tasks learning pseudo-labels, and the trained models are further fine-tuned for downstream tasks. However, one important disadvantage of SSL using pre-text task solving is defining an appropriate pre-text task for each image dataset with a variety of image modalities. Therefore, it is required to design an appropriate pretext task automatically for each dataset and each downstream task. To the best of our knowledge, the automatic designing of pretext tasks for image analysis has not been considered yet. In this paper, we present a framework based on In-context learning that describes each task based on its input and output data using a pre-trained image transformer. Our proposed method combines the input image and its learned description for optimizing the pre-text task design and its hyper-parameters using Meta-learning models. The representations learned from the pre-text tasks are fine-tuned for solving the downstream tasks. We demonstrate that our proposed framework outperforms the compared ones on unseen tasks and image modalities in addition to its superior performance for previously known tasks and datasets.

Keywords: in-context learning (ICL), meta learning, self-supervised learning (SSL), vision-language domain, transformers

Procedia PDF Downloads 81
7302 Gender Quotas in Italy: Effects on Corporate Performance

Authors: G. Bruno, A. Ciavarella, N. Linciano

Abstract:

The proportion of women in boardroom has traditionally been low around the world. Over the last decades, several jurisdictions opted for active intervention, which triggered a tangible progress in female representation. In Europe, many countries have implemented boardroom diversity policies in the form of legal quotas (Norway, Italy, France, Germany) or governance code amendments (United Kingdom, Finland). Policy actions rest, among other things, on the assumption that gender balanced boards result in improved corporate governance and performance. The investigation of the relationship between female boardroom representation and firm value is therefore key on policy grounds. The evidence gathered so far, however, has not produced conclusive results also because empirical studies on the impact of voluntary female board representation had to tackle with endogeneity, due to either differences in unobservable characteristics across firms that may affect their gender policies and governance choices, or potential reverse causality. In this paper, we study the relationship between the presence of female directors and corporate performance in Italy, where the Law 120/2011 envisaging mandatory quotas has introduced an exogenous shock in board composition which may enable to overcome reverse causality. Our sample comprises Italian firms listed on the Italian Stock Exchange and the members of their board of directors over the period 2008-2016. The study relies on two different databases, both drawn from CONSOB, referring respectively to directors and companies’ characteristics. On methodological grounds, information on directors is treated at the individual level, by matching each company with its directors every year. This allows identifying all time-invariant, possibly correlated, elements of latent heterogeneity that vary across firms and board members, such as the firm immaterial assets and the directors’ skills and commitment. Moreover, we estimate dynamic panel data specifications, so accommodating non-instantaneous adjustments of firm performance and gender diversity to institutional and economic changes. In all cases, robust inference is carried out taking into account the bidimensional clustering of observations over companies and over directors. The study shows the existence of a U-shaped impact of the percentage of women in the boardroom on profitability, as measured by Return On Equity (ROE) and Return On Assets. Female representation yields a positive impact when it exceeds a certain threshold, ranging between about 18% and 21% of the board members, depending on the specification. Given the average board size, i.e., around ten members over the time period considered, this would imply that a significant effect of gender diversity on corporate performance starts to emerge when at least two women hold a seat. This evidence supports the idea underpinning the critical mass theory, i.e., the hypothesis that women may influence.

Keywords: gender diversity, quotas, firms performance, corporate governance

Procedia PDF Downloads 171
7301 Analyzing the Street Pattern Characteristics on Young People’s Choice to Walk or Not: A Study Based on Accelerometer and Global Positioning Systems Data

Authors: Ebru Cubukcu, Gozde Eksioglu Cetintahra, Burcin Hepguzel Hatip, Mert Cubukcu

Abstract:

Obesity and overweight cause serious health problems. Public and private organizations aim to encourage walking in various ways in order to cope with the problem of obesity and overweight. This study aims to understand how the spatial characteristics of urban street pattern, connectivity and complexity influence young people’s choice to walk or not. 185 public university students in Izmir, the third largest city in Turkey, participated in the study. Each participant had worn an accelerometer and a global positioning (GPS) device for a week. The accelerometer device records data on the intensity of the participant’s activity at a specified time interval, and the GPS device on the activities’ locations. Combining the two datasets, activity maps are derived. These maps are then used to differentiate the participants’ walk trips and motor vehicle trips. Given that, the frequency of walk and motor vehicle trips are calculated at the street segment level, and the street segments are then categorized into two as ‘preferred by pedestrians’ and ‘preferred by motor vehicles’. Graph Theory-based accessibility indices are calculated to quantify the spatial characteristics of the streets in the sample. Six different indices are used: (I) edge density, (II) edge sinuosity, (III) eta index, (IV) node density, (V) order of a node, and (VI) beta index. T-tests show that the index values for the ‘preferred by pedestrians’ and ‘preferred by motor vehicles’ are significantly different. The findings indicate that the spatial characteristics of the street network have a measurable effect on young people’s choice to walk or not. Policy implications are discussed. This study is funded by the Scientific and Technological Research Council of Turkey, Project No: 116K358.

Keywords: graph theory, walkability, accessibility, street network

Procedia PDF Downloads 228
7300 Performance Comparison of Wideband Covariance Matrix Sparse Representation (W-CMSR) with Other Wideband DOA Estimation Methods

Authors: Sandeep Santosh, O. P. Sahu

Abstract:

In this paper, performance comparison of wideband covariance matrix sparse representation (W-CMSR) method with other existing wideband Direction of Arrival (DOA) estimation methods has been made.W-CMSR relies less on a priori information of the incident signal number than the ordinary subspace based methods.Consider the perturbation free covariance matrix of the wideband array output. The diagonal covariance elements are contaminated by unknown noise variance. The covariance matrix of array output is conjugate symmetric i.e its upper right triangular elements can be represented by lower left triangular ones.As the main diagonal elements are contaminated by unknown noise variance,slide over them and align the lower left triangular elements column by column to obtain a measurement vector.Simulation results for W-CMSR are compared with simulation results of other wideband DOA estimation methods like Coherent signal subspace method (CSSM), Capon, l1-SVD, and JLZA-DOA. W-CMSR separate two signals very clearly and CSSM, Capon, L1-SVD and JLZA-DOA fail to separate two signals clearly and an amount of pseudo peaks exist in the spectrum of L1-SVD.

Keywords: W-CMSR, wideband direction of arrival (DOA), covariance matrix, electrical and computer engineering

Procedia PDF Downloads 471
7299 Estimation of the Exergy-Aggregated Value Generated by a Manufacturing Process Using the Theory of the Exergetic Cost

Authors: German Osma, Gabriel Ordonez

Abstract:

The production of metal-rubber spares for vehicles is a sequential process that consists in the transformation of raw material through cutting activities and chemical and thermal treatments, which demand electricity and fossil fuels. The energy efficiency analysis for these cases is mostly focused on studying of each machine or production step, but is not common to study of the quality of the production process achieves from aggregated value viewpoint, which can be used as a quality measurement for determining of impact on the environment. In this paper, the theory of exergetic cost is used for determining of aggregated exergy to three metal-rubber spares, from an exergy analysis and thermoeconomic analysis. The manufacturing processing of these spares is based into batch production technique, and therefore is proposed the use of this theory for discontinuous flows from of single models of workstations; subsequently, the complete exergy model of each product is built using flowcharts. These models are a representation of exergy flows between components into the machines according to electrical, mechanical and/or thermal expressions; they determine the demanded exergy to produce the effective transformation in raw materials (aggregated exergy value), the exergy losses caused by equipment and irreversibilities. The energy resources of manufacturing process are electricity and natural gas. The workstations considered are lathes, punching presses, cutters, zinc machine, chemical treatment tanks, hydraulic vulcanizing presses and rubber mixer. The thermoeconomic analysis was done by workstation and by spare; first of them describes the operation of the components of each machine and where the exergy losses are; while the second of them estimates the exergy-aggregated value for finished product and wasted feedstock. Results indicate that exergy efficiency of a mechanical workstation is between 10% and 60% while this value in the thermal workstations is less than 5%; also that each effective exergy-aggregated value is one-thirtieth of total exergy required for operation of manufacturing process, which amounts approximately to 2 MJ. These troubles are caused mainly by technical limitations of machines, oversizing of metal feedstock that demands more mechanical transformation work, and low thermal insulation of chemical treatment tanks and hydraulic vulcanizing presses. From established information, in this case, it is possible to appreciate the usefulness of theory of exergetic cost for analyzing of aggregated value in manufacturing processes.

Keywords: exergy-aggregated value, exergy efficiency, thermoeconomics, exergy modeling

Procedia PDF Downloads 171
7298 Repeatable Scalable Business Models: Can Innovation Drive an Entrepreneurs Un-Validated Business Model?

Authors: Paul Ojeaga

Abstract:

Can the level of innovation use drive un-validated business models across regions? To what extent does industrial sector attractiveness drive firm’s success across regions at the time of start-up? This study examines the role of innovation on start-up success in six regions of the world (namely Sub Saharan Africa, the Middle East and North Africa, Latin America, South East Asia Pacific, the European Union and the United States representing North America) using macroeconomic variables. While there have been studies using firm level data, results from such studies are not suitable for national policy decisions. The need to drive a regional innovation policy also begs for an answer, therefore providing room for this study. Results using dynamic panel estimation show that innovation counts in the early infancy stage of new business life cycle. The results are robust even after controlling for time fixed effects and the study present variance-covariance estimation robust standard errors.

Keywords: industrial economics, un-validated business models, scalable models, entrepreneurship

Procedia PDF Downloads 283
7297 Current Methods for Drug Property Prediction in the Real World

Authors: Jacob Green, Cecilia Cabrera, Maximilian Jakobs, Andrea Dimitracopoulos, Mark van der Wilk, Ryan Greenhalgh

Abstract:

Predicting drug properties is key in drug discovery to enable de-risking of assets before expensive clinical trials and to find highly active compounds faster. Interest from the machine learning community has led to the release of a variety of benchmark datasets and proposed methods. However, it remains unclear for practitioners which method or approach is most suitable, as different papers benchmark on different datasets and methods, leading to varying conclusions that are not easily compared. Our large-scale empirical study links together numerous earlier works on different datasets and methods, thus offering a comprehensive overview of the existing property classes, datasets, and their interactions with different methods. We emphasise the importance of uncertainty quantification and the time and, therefore, cost of applying these methods in the drug development decision-making cycle. To the best of the author's knowledge, it has been observed that the optimal approach varies depending on the dataset and that engineered features with classical machine learning methods often outperform deep learning. Specifically, QSAR datasets are typically best analysed with classical methods such as Gaussian Processes, while ADMET datasets are sometimes better described by Trees or deep learning methods such as Graph Neural Networks or language models. Our work highlights that practitioners do not yet have a straightforward, black-box procedure to rely on and sets a precedent for creating practitioner-relevant benchmarks. Deep learning approaches must be proven on these benchmarks to become the practical method of choice in drug property prediction.

Keywords: activity (QSAR), ADMET, classical methods, drug property prediction, empirical study, machine learning

Procedia PDF Downloads 83
7296 The Politics of Cinema: Representation of Rising Nationalism in Indian Cinema in the Election Year of 2019

Authors: Paawani Tewari, Oishik Dasgupta

Abstract:

Cinema and politics have often intertwined in India. Movies have become a mainstream method of communication with audiences and voters subliminally and directly. Indian film industry on average produces over a thousand films in a year, and during the election year of 2019, India witnessed the release of several highly political movies. Movies such as Uri: The Surgical Strike, Accidental Prime Minister, and PM Modi, et cetera, which are the sample of this study have tried to depict an ideal character of political stalwart leaders with the plausibility to inspire and aiming to change ideological orientations of viewers and the potent voters. This study tries to understand the major links between nationalism, its representation, and its manifestation in Indian cinema and how it is instrumental in shaping the character and orientations of its citizens towards nation, nationalism, and nationhood. Our work aims to highlight how nationalistic assumptions that are swaddled in the Hindi movies released during January 2019 – May 2019 affect the political mood of the nation and, in totality, the democratic system. The authors also try to throw light on how films being a powerful tool, are now being used to shape ideas, brainwashing and swaying opinions for political mileage. Hence it becomes essential for us to explore the dynamics between the quintessential definitions of what nationalism is for a common man in India versus of what has been represented in cinema, especially during the time of the elections.

Keywords: political governance and political analysis, political and public administration, election, public choice

Procedia PDF Downloads 167
7295 Adapted Intersection over Union: A Generalized Metric for Evaluating Unsupervised Classification Models

Authors: Prajwal Prakash Vasisht, Sharath Rajamurthy, Nishanth Dara

Abstract:

In a supervised machine learning approach, metrics such as precision, accuracy, and coverage can be calculated using ground truth labels to help in model tuning, evaluation, and selection. In an unsupervised setting, however, where the data has no ground truth, there are few interpretable metrics that can guide us to do the same. Our approach creates a framework to adapt the Intersection over Union metric, referred to as Adapted IoU, usually used to evaluate supervised learning models, into the unsupervised domain, which solves the problem by factoring in subject matter expertise and intuition about the ideal output from the model. This metric essentially provides a scale that allows us to compare the performance across numerous unsupervised models or tune hyper-parameters and compare different versions of the same model.

Keywords: general metric, unsupervised learning, classification, intersection over union

Procedia PDF Downloads 50
7294 Accelerating Molecular Dynamics Simulations of Electrolytes with Neural Network: Bridging the Gap between Ab Initio Molecular Dynamics and Classical Molecular Dynamics

Authors: Po-Ting Chen, Santhanamoorthi Nachimuthu, Jyh-Chiang Jiang

Abstract:

Classical molecular dynamics (CMD) simulations are highly efficient for material simulations but have limited accuracy. In contrast, ab initio molecular dynamics (AIMD) provides high precision by solving the Kohn–Sham equations yet requires significant computational resources, restricting the size of systems and time scales that can be simulated. To address these challenges, we employed NequIP, a machine learning model based on an E(3)-equivariant graph neural network, to accelerate molecular dynamics simulations of a 1M LiPF6 in EC/EMC (v/v 3:7) for Li battery applications. AIMD calculations were initially conducted using the Vienna Ab initio Simulation Package (VASP) to generate highly accurate atomic positions, forces, and energies. This data was then used to train the NequIP model, which efficiently learns from the provided data. NequIP achieved AIMD-level accuracy with significantly less training data. After training, NequIP was integrated into the LAMMPS software to enable molecular dynamics simulations of larger systems over longer time scales. This method overcomes the computational limitations of AIMD while improving the accuracy limitations of CMD, providing an efficient and precise computational framework. This study showcases NequIP’s applicability to electrolyte systems, particularly for simulating the dynamics of LiPF6 ionic mixtures. The results demonstrate substantial improvements in both computational efficiency and simulation accuracy, highlighting the potential of machine learning models to enhance molecular dynamics simulations.

Keywords: lithium-ion batteries, electrolyte simulation, molecular dynamics, neural network

Procedia PDF Downloads 25
7293 Literature Review and Approach for the Use of Digital Factory Models in an Augmented Reality Application for Decision Making in Restructuring Processes

Authors: Rene Hellmuth, Jorg Frohnmayer

Abstract:

The requirements of the factory planning and the building concerned have changed in the last years. Factory planning has the task of designing products, plants, processes, organization, areas, and the building of a factory. Regular restructuring gains more importance in order to maintain the competitiveness of a factory. Even today, the methods and process models used in factory planning are predominantly based on the classical planning principles of Schmigalla, Aggteleky and Kettner, which, however, are not specifically designed for reorganization. In addition, they are designed for a largely static environmental situation and a manageable planning complexity as well as for medium to long-term planning cycles with a low variability of the factory. Existing approaches already regard factory planning as a continuous process that makes it possible to react quickly to adaptation requirements. However, digital factory models are not yet used as a source of information for building data. Approaches which consider building information modeling (BIM) or digital factory models in general either do not refer to factory conversions or do not yet go beyond a concept. This deficit can be further substantiated. A method for factory conversion planning using a current digital building model is lacking. A corresponding approach must take into account both the existing approaches to factory planning and the use of digital factory models in practice. A literature review will be conducted first. In it, approaches to classic factory planning and approaches to conversion planning are examined. In addition, it will be investigated which approaches already contain digital factory models. In the second step, an approach is presented how digital factory models based on building information modeling can be used as a basis for augmented reality tablet applications. This application is suitable for construction sites and provides information on the costs and time required for conversion variants. Thus a fast decision making is supported. In summary, the paper provides an overview of existing factory planning approaches and critically examines the use of digital tools. Based on this preliminary work, an approach is presented, which suggests the sensible use of digital factory models for decision support in the case of conversion variants of the factory building. The augmented reality application is designed to summarize the most important information for decision-makers during a reconstruction process.

Keywords: augmented reality, digital factory model, factory planning, restructuring

Procedia PDF Downloads 138
7292 UniFi: Universal Filter Model for Image Enhancement

Authors: Aleksei Samarin, Artyom Nazarenko, Valentin Malykh

Abstract:

Image enhancement is becoming more and more popular, especially on mobile devices. Nowadays, it is a common approach to enhance an image using a convolutional neural network (CNN). Such a network should be of significant size; otherwise, a possibility for the artifacts to occur is overgrowing. The existing large CNNs are computationally expensive, which could be crucial for mobile devices. Another important flaw of such models is they are poorly interpretable. There is another approach to image enhancement, namely, the usage of predefined filters in combination with the prediction of their applicability. We present an approach following this paradigm, which outperforms both existing CNN-based and filter-based approaches in the image enhancement task. It is easily adaptable for mobile devices since it has only 47 thousand parameters. It shows the best SSIM 0.919 on RANDOM250 (MIT Adobe FiveK) among small models and is thrice faster than previous models.

Keywords: universal filter, image enhancement, neural networks, computer vision

Procedia PDF Downloads 102
7291 Smart Web Services in the Web of Things

Authors: Sekkal Nawel

Abstract:

The Web of Things (WoT), integration of smart technologies from the Internet or network to Web architecture or application, is becoming more complex, larger, and dynamic. The WoT is associated with various elements such as sensors, devices, networks, protocols, data, functionalities, and architectures to perform services for stakeholders. These services operate in the context of the interaction of stakeholders and the WoT elements. Such context is becoming a key information source from which data are of various nature and uncertain, thus leading to complex situations. In this paper, we take interest in the development of intelligent Web services. The key ingredients of this “intelligent” notion are the context diversity, the necessity of a semantic representation to manage complex situations and the capacity to reason with uncertain data. In this perspective, we introduce a multi-layered architecture based on a generic intelligent Web service model dealing with various contexts, which proactively predict future situations and reactively respond to real-time situations in order to support decision-making. For semantic context data representation, we use PR-OWL, which is a probabilistic ontology based on Multi-Entity Bayesian Networks (MEBN). PR-OWL is flexible enough to represent complex, dynamic, and uncertain contexts, the key requirements of the development for the intelligent Web services. A case study was carried out using the proposed architecture for intelligent plant watering to show the role of proactive and reactive contextual reasoning in terms of WoT.

Keywords: smart web service, the web of things, context reasoning, proactive, reactive, multi-entity bayesian networks, PR-OWL

Procedia PDF Downloads 72
7290 Characteristics of Inclusive Circular Business Models in Social Entrepreneurship

Authors: Svitlana Yermak, Olubukola Aluko

Abstract:

The purpose of this study was a literature review on the topic of social entrepreneurship, a review of new trends and best practices, the study of existing inclusive business models and their interaction with the principles of the circular economy for possible implementation in the practice of Ukraine in war and post-war times in conditions of scarce resources. Thus, three research questions were identified and substantiated: to determine the characteristics of social entrepreneurship, consider the features in Ukraine and the UK; highlight the criteria for inclusion in social entrepreneurship and its legal support; explore examples of existing inclusive circular business models to illustrate how the two concepts may be combined. A detailed review of the literature selected from the Scopus and Web of Science databases was carried out. The study revealed signs of social entrepreneurship, the main of which are doing business and making a profit, as well as the social orientation of the business, which is prescribed in the constituent documents of the enterprise immediately upon its creation. Considered are the characteristics of social entrepreneurship in the UK and Ukraine. It has been established that in the UK, social entrepreneurship is clearly regulated by the state; there are special legislative norms and support programs, in contrast to Ukraine, where these processes are only partially regulated. The study identified the main criteria for inclusion in inclusive circular business models: economic (sustainability and efficiency, job creation and economic growth, promotion of local development), social (accessibility, equity and fairness, inclusion and participation), and resources in their interconnection. It is substantiated that the resource criterion is especially important for this type of business model. It provides for the efficient and sustainable use of resources, as well as the cyclical nature of resources. And it was concluded that the principles of the circular economy not only do not contradict but, on the contrary, complement and expand the inclusive business models on which social entrepreneurship is based.

Keywords: social entrepreneurship, inclusive business models, circular economy, inclusion criteria

Procedia PDF Downloads 102
7289 Comparison of Machine Learning Models for the Prediction of System Marginal Price of Greek Energy Market

Authors: Ioannis P. Panapakidis, Marios N. Moschakis

Abstract:

The Greek Energy Market is structured as a mandatory pool where the producers make their bid offers in day-ahead basis. The System Operator solves an optimization routine aiming at the minimization of the cost of produced electricity. The solution of the optimization problem leads to the calculation of the System Marginal Price (SMP). Accurate forecasts of the SMP can lead to increased profits and more efficient portfolio management from the producer`s perspective. Aim of this study is to provide a comparative analysis of various machine learning models such as artificial neural networks and neuro-fuzzy models for the prediction of the SMP of the Greek market. Machine learning algorithms are favored in predictions problems since they can capture and simulate the volatilities of complex time series.

Keywords: deregulated energy market, forecasting, machine learning, system marginal price

Procedia PDF Downloads 216
7288 Between Reality and Fiction: Self-Representation as an Avatar and Its Effects on Self-Presence

Authors: Leonie Laskowitz

Abstract:

A self-confident appearance is a basic prerequisite for success in the world of work 4.0. Within a few seconds, people convey a first impression that usually lasts. Artificial intelligence is making it increasingly important how our virtual selves appear and communicate (nonverbally) in digital worlds such as the metaverse. In addition to the modified creation of an avatar, the field of photogrammetry is developing fast, creating exact likenesses of ourselves in virtual environments. Given the importance of self-representation in virtual space for future collaborations, it is important to investigate the impact of phenotype in virtual worlds and how an avatar type can profitably be used situationally. We analyzed the effect of self-similar versus desirable self-presentation as an avatar on one's self-awareness, considering various theoretical constructs in the area of self-awareness and stress stimuli. The avatars were arbitrarily created on the one hand and scanned on the other hand with the help of a lidar sensor, the state-of-the-art photogrammetry method. All subjects were exposed to the established Trier Social Stress Test. The results showed that especially insecure people prefer to create rather than be scanned when confronted with a stressful work situation. (1) If they are in a casual work environment and a relaxed situation, they prefer a 3D photorealistic avatar that reflects them in detail. (2) Confident people will give their avatar their true appearance in any situation, while insecure people would only do so for honesty and authenticity. (3) Thus, the choice of avatar type has considerable impact on self-confidence in different situations.

Keywords: avatar, virtual identity, self-presentation, metaverse, virtual reality, self-awareness

Procedia PDF Downloads 149
7287 An investigation of the High-frequency Isolation Performance of Quasi-Zero-Stiffness Vibration Isolators

Authors: Chen Zhang, Yongpeng Gu, Xiaotian Li

Abstract:

Quasi-zero-stiffness (QZS) vibration isolation technology has garnered significant attention in both academia and industry, which enables ultra-low-frequency vibration isolation. In modern industries, such as shipbuilding and aerospace, rotating machinery generates vibrations over a wide frequency range, thus imposing more stringent requirements on vibration isolation technologies. These technologies must not only achieve ultra-low starting isolation frequencies but also provide effective isolation across mid- to high-frequency ranges. However, existing research on QZS vibration isolators primarily focuses on frequency ranges below 50 Hz. Moreover, studies have shown that in the mid-to high-frequency ranges, QZS isolators tend to generate resonance peaks that adversely affect their isolation performance. This limitation significantly restricts the practical applicability of QZS isolation technology. To address this issue, the present study investigates the high-frequency isolation performance of two typical QZS isolators: the mechanism type three-spring QZS isolator mechanism and the structure and bowl-shaped QZS isolator structure. First, the parameter conditions required to achieve quasi-zero stiffness characteristics for two isolators are derived based on static mechanical analysis. The theoretical transmissibility characteristics are then calculated using the harmonic balance method. Three-dimensional finite element models of two QZS isolators are developed using ABAQUS simulation software, and transmissibility curves are computed for the 0-500 Hz frequency range. The results indicate that the three-spring QZS mechanism exhibits multiple higher-order resonance peaks in the mid-to high-frequency ranges due to the higher-order models of the springs. Springs with fewer coils and larger diameters can shift the higher-order modals to higher frequencies but cannot entirely eliminate their occurrence. In contrast, the bowl-shaped QZS isolator, through shape optimization using a spline-based representation, effectively mitigates the generation of higher-order resonance modes, resulting in superior isolation performance in the mid-to high-frequency ranges. This study provides essential theoretical insights for optimizing the vibration isolation performance of QZS technologies in complex, wide-frequency vibration environments, offering significant practical value for their application.

Keywords: quasi-zero-stiffness, wide-frequency vibration, vibration isolator, transmissibility

Procedia PDF Downloads 12
7286 Improving Cell Type Identification of Single Cell Data by Iterative Graph-Based Noise Filtering

Authors: Annika Stechemesser, Rachel Pounds, Emma Lucas, Chris Dawson, Julia Lipecki, Pavle Vrljicak, Jan Brosens, Sean Kehoe, Jason Yap, Lawrence Young, Sascha Ott

Abstract:

Advances in technology make it now possible to retrieve the genetic information of thousands of single cancerous cells. One of the key challenges in single cell analysis of cancerous tissue is to determine the number of different cell types and their characteristic genes within the sample to better understand the tumors and their reaction to different treatments. For this analysis to be possible, it is crucial to filter out background noise as it can severely blur the downstream analysis and give misleading results. In-depth analysis of the state-of-the-art filtering methods for single cell data showed that they do, in some cases, not separate noisy and normal cells sufficiently. We introduced an algorithm that filters and clusters single cell data simultaneously without relying on certain genes or thresholds chosen by eye. It detects communities in a Shared Nearest Neighbor similarity network, which captures the similarities and dissimilarities of the cells by optimizing the modularity and then identifies and removes vertices with a weak clustering belonging. This strategy is based on the fact that noisy data instances are very likely to be similar to true cell types but do not match any of these wells. Once the clustering is complete, we apply a set of evaluation metrics on the cluster level and accept or reject clusters based on the outcome. The performance of our algorithm was tested on three datasets and led to convincing results. We were able to replicate the results on a Peripheral Blood Mononuclear Cells dataset. Furthermore, we applied the algorithm to two samples of ovarian cancer from the same patient before and after chemotherapy. Comparing the standard approach to our algorithm, we found a hidden cell type in the ovarian postchemotherapy data with interesting marker genes that are potentially relevant for medical research.

Keywords: cancer research, graph theory, machine learning, single cell analysis

Procedia PDF Downloads 114
7285 Dynamic Modeling of Advanced Wastewater Treatment Plants Using BioWin

Authors: Komal Rathore, Aydin Sunol, Gita Iranipour, Luke Mulford

Abstract:

Advanced wastewater treatment plants have complex biological kinetics, time variant influent flow rates and long processing times. Due to these factors, the modeling and operational control of advanced wastewater treatment plants become complicated. However, development of a robust model for advanced wastewater treatment plants has become necessary in order to increase the efficiency of the plants, reduce energy costs and meet the discharge limits set by the government. A dynamic model was designed using the Envirosim (Canada) platform software called BioWin for several wastewater treatment plants in Hillsborough County, Florida. Proper control strategies for various parameters such as mixed liquor suspended solids, recycle activated sludge and waste activated sludge were developed for models to match the plant performance. The models were tuned using both the influent and effluent data from the plant and their laboratories. The plant SCADA was used to predict the influent wastewater rates and concentration profiles as a function of time. The kinetic parameters were tuned based on sensitivity analysis and trial and error methods. The dynamic models were validated by using experimental data for influent and effluent parameters. The dissolved oxygen measurements were taken to validate the model by coupling them with Computational Fluid Dynamics (CFD) models. The Biowin models were able to exactly mimic the plant performance and predict effluent behavior for extended periods. The models are useful for plant engineers and operators as they can take decisions beforehand by predicting the plant performance with the use of BioWin models. One of the important findings from the model was the effects of recycle and wastage ratios on the mixed liquor suspended solids. The model was also useful in determining the significant kinetic parameters for biological wastewater treatment systems.

Keywords: BioWin, kinetic modeling, flowsheet simulation, dynamic modeling

Procedia PDF Downloads 155
7284 Mission Driven Enterprises in Ecosystems as Drivers for Sustainable System Change

Authors: Monique de Ritter, Annemieke Roobeek

Abstract:

This study takes a holistic multi-layered systems approach on entrepreneurship, innovation, and sustainability. Concretely we looked how mission driven entrepreneurs (level 1) employ new business models and launch innovative products and/or ideas in their enterprises, which are (level 2) operating in entrepreneurial ecosystems (level 3), and how these in turn may generate higher level sustainable change (level 4). We employed a qualitative grounded research approach in which our aim is to contribute to theory. Fourteen in-depth semi-structured interviews were conducted with mission driven entrepreneurs in the Netherlands in which their individual drives, business models, and ecosystems were discussed. Interview transcripts were systematically coded and analysed and the ecosystems were visually mapped. Most important patterns include 1) entrepreneurs have a clear sustainable mission and regard this mission as de raison d’être of their enterprise; 2) entrepreneurs employ new business models with a focus on collaboration for innovation; the business model supports or enhances the sustainable mission of the enterprise, 3) entrepreneurs collaborate in ecosystems in which a) they also regard suppliers as partners for innovation and clients as ambassadors for the sustainable mission, b) would like to improve their relationships with financial institutions as they are in the entrepreneurs’ perspective often lagging behind with their innovative ideas and models, c) they collaborate for knowledge and innovation with several parties, d) personal informal connections are very important, and e) in which the higher sustainable mission is not a point of competition but of collaboration.

Keywords: sustainability, entrepreneurship, innovation, ecosystem, business models

Procedia PDF Downloads 376