Search results for: hybrid models

Authors: Armineh Nourbakhsh, Sameena Shah, Carolyn Rose

Abstract:

Layout-aware language models have been used to create multimodal representations for documents that are in image form, achieving relatively high accuracy in document understanding tasks. However, the large number of parameters in the resulting models makes building and using them prohibitive without access to high-performing processing units with large memory capacity. We propose an alternative approach that can create efficient representations without the need for a neural visual backbone. This leads to an 80% reduction in the number of parameters compared to the smallest SOTA model, widely expanding applicability. In addition, our layout embeddings are pre-trained on spatial and visual cues alone and only fused with text embeddings in downstream tasks, which can facilitate applicability to low-resource of multi-lingual domains. Despite using 2.5% of training data, we show competitive performance on two form understanding tasks: semantic labeling and link prediction.

Keywords: layout understanding, form understanding, multimodal document understanding, bias-augmented attention

Procedia PDF Downloads 148

Authors: Yepeng Cheng, Yasuhiko Morimoto

Abstract:

Customer relationship analysis is vital for retail stores, especially for supermarkets. The point of sale (POS) systems make it possible to record the daily purchasing behaviors of customers as an identification point of sale (ID-POS) database, which can be used to analyze customer behaviors of a supermarket. The customer value is an indicator based on ID-POS database for detecting the customer loyalty of a store. In general, there are many supermarkets in a city, and other nearby competitor supermarkets significantly affect the customer value of customers of a supermarket. However, it is impossible to get detailed ID-POS databases of competitor supermarkets. This study firstly focused on the customer value and distance between a customer's home and supermarkets in a city, and then constructed the models based on logistic regression analysis to analyze correlations between distance and purchasing behaviors only from a POS database of a supermarket chain. During the modeling process, there are three primary problems existed, including the incomparable problem of customer values, the multicollinearity problem among customer value and distance data, and the number of valid partial regression coefficients. The improved customer value, Huff’s gravity model, and inverse attractiveness frequency are considered to solve these problems. This paper presents three types of models based on these three methods for loyal customer classification and competitors’ influence analysis. In numerical experiments, all types of models are useful for loyal customer classification. The type of model, including all three methods, is the most superior one for evaluating the influence of the other nearby supermarkets on customers' purchasing of a supermarket chain from the viewpoint of valid partial regression coefficients and accuracy.

Keywords: customer value, Huff's Gravity Model, POS, Retailer

Procedia PDF Downloads 123

Authors: Nasim Eslamirad, Mahdi Rasoulinezhad, Francesco De Luca, Sadok Ben Yahia, Kimmo Sakari Lylykangas, Francesco Pilla

Abstract:

With the ongoing urbanization, cities face increasing environmental challenges impacting human well-being. To tackle these issues, data-driven approaches in urban analysis have gained prominence, leveraging urban data to promote sustainability. Integrating Machine Learning techniques enables researchers to analyze and predict complex environmental phenomena like Urban Heat Island occurrences in urban areas. This paper demonstrates the implementation of data-driven approach and interpretable Machine Learning algorithms with interpretability techniques to conduct comprehensive data analyses for sustainable urban design. The developed framework and algorithms are demonstrated for Tallinn, Estonia to develop sustainable urban strategies to mitigate urban heat waves. Geospatial data, preprocessed and labeled with UHI levels, are used to train various ML models, with Logistic Regression emerging as the best-performing model based on evaluation metrics to derive a mathematical equation representing the area with UHI or without UHI effects, providing insights into UHI occurrences based on buildings and urban features. The derived formula highlights the importance of building volume, height, area, and shape length to create an urban environment with UHI impact. The data-driven approach and derived equation inform mitigation strategies and sustainable urban development in Tallinn and offer valuable guidance for other locations with varying climates.

Keywords: data-driven approach, machine learning transparent models, interpretable machine learning models, urban heat island effect

Procedia PDF Downloads 37

Authors: Mais Khader, Xingjie Wei

Abstract:

This research sheds light on female entrepreneurs by providing new insights on the survival predictions of companies led by females in the UK. This study aims to build a predictive machine learning model of the survival of female-led & co-led small & medium enterprises (SMEs) in the UK over the period 2000-2020. The predictive model built utilised a combination of financial and non-financial features related to both companies and their directors to predict SMEs' survival. These features were studied in terms of their contribution to the resultant predictive model. Five machine learning models are used in the modelling: Decision tree, AdaBoost, Naïve Bayes, Logistic regression and SVM. The AdaBoost model had the highest performance of the five models, with an accuracy of 73% and an AUC of 80%. The results show high feature importance in predicting companies' survival for company size, management experience, financial performance, industry, region, and females' percentage in management.

Keywords: company survival, entrepreneurship, females, machine learning, SMEs

Procedia PDF Downloads 101

Authors: Ya-Li Tsai

Abstract:

Nitrogen is a vital element crucial for crop growth, significantly influencing crop yield. In rice cultivation, farmers often apply substantial nitrogen fertilizer to maximize yields. However, excessive nitrogen application increases the risk of lodging and pest infestation, leading to yield losses. Additionally, conventional flooded irrigation methods consume significant water resources, necessitating precise agricultural and intelligent water management systems. In this study, it leveraged physiological data and field images captured by unmanned aerial vehicles, considering fertilizer treatment and irrigation as key factors. Statistical models incorporating rice physiological data, yield, and vegetation indices from image data were developed. Missing physiological data were addressed using multiple imputation and regression methods, and regression models were established using principal component analysis and stepwise regression. Target nitrogen accumulation at key growth stages was identified to optimize fertilizer application, with the difference between actual and target nitrogen accumulation guiding recommendations for ear dressing dosage. Field experiments conducted in 2022 validated the recommended ear dressing dosage, demonstrating no significant difference in final yield compared to traditional fertilizer levels under alternate wetting and drying irrigation. These findings highlight the efficacy of applying recommended dosages based on fertilizer decision models, offering the potential for reduced fertilizer use while maintaining yield in rice cultivation.

Keywords: intelligent fertilizer management, nitrogen top and ear dressing fertilizer, rice, yield optimization

Procedia PDF Downloads 82

Authors: Amro Issam Hamed Attia Ramadan, Marco Zappatore, Pasquale Balena, Antonella Longo

Abstract:

Considering the energetic crisis that is hitting Europe, it becomes more and more necessary to change the energy policies to depend less on fossil fuels and replace them with energy from renewable sources. This has triggered the urge to use clean energy not only to satisfy energy needs and fulfill the required consumption but also to decrease the danger of climatic changes due to harmful emissions. Many countries have already started creating energetic communities based on renewable energy sources. The first step to understanding energy needs in any place is to perfectly know the consumption. In this work, we aim to estimate electricity consumption for a municipality that makes up part of a rural area located in southern Italy using forecast models that allow for the estimation of electricity consumption for the next ten years, and we then apply the same model to the province where the municipality is located and estimate the future consumption for the same period to examine whether it is possible to start from the municipality level to reach the province level when creating energy communities.

Keywords: ARIMA, electricity consumption, forecasting models, time series

Procedia PDF Downloads 174

Authors: Martin Cermak, Tomas Karasek, Jaroslav Rojicek

Abstract:

The key role in phenomenological modelling of cyclic plasticity is good understanding of stress-strain behaviour of given material. There are many models describing behaviour of materials using numerous parameters and constants. Combination of individual parameters in those material models significantly determines whether observed and predicted results are in compliance. Parameter identification techniques such as random gradient, genetic algorithm, and sensitivity analysis are used for identification of parameters using numerical modelling and simulation. In this paper genetic algorithm and sensitivity analysis are used to study effect of 4 parameters of modified AbdelKarim-Ohno cyclic plasticity model. Results predicted by Finite Element (FE) simulation are compared with experimental data from biaxial ratcheting test with semi-elliptical loading path.

Keywords: genetic algorithm, sensitivity analysis, inverse approach, finite element method, cyclic plasticity, ratcheting

Procedia PDF Downloads 453

Authors: Jocelyne Kenne Kenne

Abstract:

Previous works in the field of sociolinguistics have explored the various outcomes of linguistic pluralism. One of these outcomes is the creation of new languages. The presentation will focus on one of such languages, Camfranglais, a hybrid language spoken by Cameroonians. It appeared in the 1970s in the francophone area in Cameroon and developed as a result of interactions between French, English, Cameroonian Pidgin English and local Cameroonian languages, all languages spoken in Cameroon. With the migration of Cameroonians to Europe, researches have been conducted to analyze the sociolinguistic profile of Cameroonians in their new environment. The emphasis on this presentation will be on two recent studies that have been conducted to analyze the peculiarity of Camfranglais in two European countries: Germany and Italy. The research involved 59 Cameroonians living in Italy and 49 Cameroonians residing in Germany. The respondents were composed of participants from different linguistic background, students and workers, married and single. A combination of quantitative and qualitative research methods was employed. The field study was divided into three parts. The first part was focused on observing the Cameroonians interact in different places such as in canteens, in the university halls of residence, lecture theatres, at homes, at various Cameroonian meetings. Those observations were accompanied by audio-recordings of the various interactions. The aim was to study communication between Cameroonians to see whether they use Camfranglais or not; if yes, in which domains and what were the speakers’ linguistic profiles. Additionally, questionnaires of different lengths were used to collect biographical information concerning the participants and their sociolinguistic profile and finally, in-depth interviews with Cameroonians were conducted to inquire about the use, the functions and the importance of this language in the migratory context. The results of the research demonstrate how a widespread use of Camfranglais by Cameroonians in Germany and Italy reveal a longing for home on the one hand and a sign of belonging on the other. It also shows the differences that exist between the profiles of Camfranglais speakers in Europe and the speakers in Cameroon notably in terms of age and social class. Finally, it points out some differences in the use, the structure and the functions of this hybrid language in the migratory setting. This study is a contribution to existing research in the field of contact languages and can serve as a comparison for other situations of multilingualism and the creation of mixed languages. Furthermore, with globalization, the study of migrant languages and the contact of these languages with new languages are topics that might be productive for further research in the field of sociolinguistics.

Keywords: interaction, migrants language, multilingualism, mixed languages

Procedia PDF Downloads 213

Authors: Mohammad A. Al-Madi, Nagham A. Al-Madi, Fanan A. Al-Madi

Abstract:

The current concept of the technology of cloud computing libraries has been increasing where users can store their data in a virtual space and can be retrieved from anywhere whilst using the network. By using cloud computing technology, industries and individuals save money, time, and space. Moreover, data and information about libraries can be placed in the cloud. This paper discusses the meaning of cloud computing along with its types. Further, the focus has been given to the application of cloud computing in modern libraries. Additionally, the advantages of cloud computing and the areas in which cloud computing be applied with current usage are discussed. Finally, the present situation of the Jordanian libraries is considered and discussed in further detail.

Keywords: cloud computing, community cloud, hybrid cloud, private cloud, public cloud

Procedia PDF Downloads 221

Authors: Alejandra Sierra, Gloria Valadez, Adriana Dávalos, Mirliana Ramírez

Abstract:

For years, the Pan American Health Organization/World Health Organization and other organizations have made efforts to the improve access and the quality of prenatal care as part of comprehensive programs for maternal and neonatal health, the standards of care have been renewed in order to migrate from a medical perspective to a holistic perspective. However, despite the efforts currently antenatal care models have not been verified by a scientific evaluation in order to determine their effectiveness. The teenage pregnancy is considered as a very important phenomenon since it has been strongly associated with inequalities, poverty and the lack of gender quality; therefore it is important to analyze the antenatal care that’s been given, including not only the clinical intervention but also the activities surrounding the advertising and the health education. In this study, the objective was to describe if the previously established activities (on the prenatal care models) are being performed in the care of pregnant teenagers attending prenatal care in health institutions in two cities in México and Chile during 2013. Methods: Observational and descriptive study, of a transversal cohort. 170 pregnant women (13-19 years) were included in prenatal care in two health institutions (100 women from León-Mexico and 70 from Chile-Coquimbo). Data collection: direct survey, perinatal clinical record card which was used as checklists: WHO antenatal care model WHO-2003, Official Mexican Standard NOM-007-SSA2-1993 and Personalized Service Manual on Reproductive Process- Chile Crece Contigo; for data analysis descriptive statistics were used. The project was approved by the relevant ethics committees. Results: Regarding the fulfillment of interventions focused on physical, gynecological exam, immunizations, monitoring signs and biochemical parameters in both groups was met by more than 84%; the activities of guidance and counseling pregnant teenagers in Leon compliance rates were below 50%, on the other hand, although pregnant women in Coquimbo had a higher percentage of compliance, no one reached 100%. The topics that less was oriented were: family planning, signs and symptoms of complications and labor. Conclusions: Although the coverage of the interventions indicated in the prenatal care models was high, there were still shortcomings in the fulfillment of activities to orientation, education and health promotion. Deficiencies in adherence to prenatal care guidelines could be due to different circumstances such as lack of registration or incomplete filling of medical records, lack of medical supplies or health personnel, absences of people at prenatal check-up appointments, among many others. Therefore, studies are required to evaluate the quality of prenatal care and the effectiveness of existing models, considering the role of the different actors (pregnant women, professionals and health institutions) involved in the functionality and quality of prenatal care models, in order to create strategies to design or improve the application of a complete process of promotion and prevention of maternal and child health as well as sexual and reproductive health in general.

Keywords: adolescent health, health systems, maternal health, primary health care

Procedia PDF Downloads 206

Authors: Owamah Hilary, O. C. Izinyon

Abstract:

Many existing biogas kinetic models are difficult to apply to substrates they were not developed for, as they are substrate specific. Biodegradability kinetic (BIK) model and maximum biogas production potential and stability assessment (MBPPSA) model were therefore developed in this study for the anaerobic co-digestion of food waste and maize husk. Biodegradability constant (k) was estimated as 0.11d-1 using the BIK model. The results of maximum biogas production potential (A) obtained using the MBPPSA model corresponded well with the results obtained using the popular but complex modified Gompertz model for digesters B-1, B-2, B-3, B-4, and B-5. The (If) value of MBPPSA model also showed that digesters B-3, B-4, and B-5 were stable, while B-1 and B-2 were unstable. Similar stability observation was also obtained using the modified Gompertz model. The MBPPSA model can therefore be used as alternative model for anaerobic digestion feasibility studies and plant design.

Keywords: biogas, inoculum, model development, stability assessment

Procedia PDF Downloads 429

Authors: Buddhi Arachchige, Hessam Ghasemnejad

Abstract:

In this study, the post impact behavior of curved composite plates subjected to low velocity impact was studied analytically and numerically. Approaches to damage modelling are proposed through the degradation of stiffness in the damaged region by reduction of thickness in the damage region. Spring-mass models were used to model the impact response of the plate and impactor. The study involved designing two damage models to compare and contrast the model best fitted with the numerical results. The theoretical force-time responses were compared with the numerical results obtained through a detailed study carried out in LS-DYNA. The modified damage model established a good prediction with the analytical force-time response for different layups and geometry. This study provides a gateway in selecting the most effective layups for variable stiffness curved composite panels able to withstand a higher impact damage.

Keywords: analytical modelling, composite damage, impact, variable stiffness

Procedia PDF Downloads 277

Authors: H. J. Joshi, Satyajeet Patil, Parth Dandavate, Mihir Kulkarni, Harshita Agrawal

Abstract:

As the world looks towards a sustainable future, electric vehicles have become increasingly popular. Millions worldwide are looking to switch to Electric cars over the previously favored combustion engine-powered cars. This demand has seen an increase in Electric Vehicle Charging Stations. The big challenge is that the randomness of electrical energy makes it tough for these charging stations to provide an adequate amount of energy over a specific amount of time. Thus, it has become increasingly crucial to model these patterns and forecast the energy needs of power stations. This paper aims to analyze how different machine learning models perform on Electric Vehicle charging time-series data. The data set consists of authentic Electric Vehicle Data from the Netherlands. It has an overview of ten thousand transactions from public stations operated by EVnetNL.

Keywords: forecasting, smart grid, electric vehicle load forecasting, machine learning, time series forecasting

Procedia PDF Downloads 106

Authors: Tin Thongthammachart, Wanida Jinsart

Abstract:

Atmospheric particulate matter particles with a diameter less than 10 microns (PM₁₀) and less than 2.5 microns (PM₂.₅) have adverse health effect. The impact from PM was studied from both health and regulatory perspective. Ambient PM data was collected over ten years in Bangkok and vicinity areas of Thailand from 2007 to 2017. Statistical models were used to forecast PM concentrations from 2018 to 2020. Monitoring monthly data averaged concentration of PM₁₀ and PM₂.₅ were used as input to forecast the monthly average concentration of PM. The forecasting results were validated by root means square error (RMSE). The predicted results were used to determine hazard risk for the carcinogenic disease. The health risk values were interpolated with GIS with ordinary kriging technique to create hazard maps in Bangkok and vicinity area. GIS-based maps illustrated the variability of PM distribution and high-risk locations. These evaluated results could support national policy for the sake of human health.

Keywords: PM₁₀, PM₂.₅, statistical models, atmospheric particulate matter

Procedia PDF Downloads 159

Authors: Dino Carpentras, Paul Maher, Caoimhe O'Reilly, Michael Quayle

Abstract:

Opinion dynamics is dedicated to modeling the dynamic evolution of people's opinions. Models in this field are based on a micro-dynamic rule, which determines how people update their opinion when interacting. Despite the high number of new models (many of them based on new rules), little research has been dedicated to experimentally validate the rule. A few studies started bridging this literature gap by experimentally testing the rule. However, in these studies, participants are forced to express their opinion as a number instead of using natural language. Furthermore, some of these studies average data from experimental questions, without testing if differences existed between them. Indeed, it is possible that different topics could show different dynamics. For example, people may be more prone to accepting someone's else opinion regarding less polarized topics. In this work, we collected data from 200 participants on 5 unpolarized topics. Participants expressed their opinions using natural language ('agree' or 'disagree') and the certainty of their answer, expressed as a number between 1 and 10. To keep the interaction based on natural language, certainty was not shown to other participants. We then showed to the participant someone else's opinion on the same topic and, after a distraction task, we repeated the measurement. To produce data compatible with standard opinion dynamics models, we multiplied the opinion (encoded as agree=1 and disagree=-1) with the certainty to obtain a single 'continuous opinion' ranging from -10 to 10. By analyzing the topics independently, we observed that each one shows a different initial distribution. However, the dynamics (i.e., the properties of the opinion change) appear to be similar between all topics. This suggested that the same micro-dynamic rule could be applied to unpolarized topics. Another important result is that participants that change opinion tend to maintain similar levels of certainty. This is in contrast with typical micro-dynamics rules, where agents move to an average point instead of directly jumping to the opposite continuous opinion. As expected, in the data, we also observed the effect of social influence. This means that exposing someone with 'agree' or 'disagree' influenced participants to respectively higher or lower values of the continuous opinion. However, we also observed random variations whose effect was stronger than the social influence’s one. We even observed cases of people that changed from 'agree' to 'disagree,' even if they were exposed to 'agree.' This phenomenon is surprising, as, in the standard literature, the strength of the noise is usually smaller than the strength of social influence. Finally, we also built an opinion dynamics model from the data. The model was able to explain more than 80% of the data variance. Furthermore, by iterating the model, we were able to produce polarized states even starting from an unpolarized population. This experimental approach offers a way to test the micro-dynamic rule. This also allows us to build models which are directly grounded on experimental results.

Keywords: experimental validation, micro-dynamic rule, opinion dynamics, update rule

Procedia PDF Downloads 162

Authors: Wendu Pang, Yaxin Luo, Junhong Li, Yu Zhao, Danni Cheng, Yufang Rao, Minzi Mao, Ke Qiu, Yijun Dong, Fei Chen, Jun Liu, Jian Zou, Haiyang Wang, Wei Xu, Jianjun Ren

Abstract:

We aimed to compare the prognostic prediction value of positive lymph node number (PLNN) to the American Joint Committee on Cancer (AJCC) tumor, lymph node, and metastasis (TNM) staging system for patients with hypopharyngeal squamous cell carcinoma (HPSCC). A total of 826 patients with HPSCC from the Surveillance, Epidemiology, and End Results database (2004–2015) were identified and split into two independent cohorts: training (n=461) and validation (n=365). Univariate and multivariate Cox regression analyses were used to evaluate the prognostic effects of PLNN in patients with HPSCC. We further applied six Cox regression models to compare the survival predictive values of the PLNN and AJCC TNM staging system. PLNN showed a significant association with overall survival (OS) and cancer-specific survival (CSS) (P < 0.001) in both univariate and multivariable analyses, and was divided into three groups (PLNN 0, PLNN 1-5, and PLNN>5). In the training cohort, multivariate analysis revealed that the increased PLNN of HPSCC gave rise to significantly poor OS and CSS after adjusting for age, sex, tumor size, and cancer stage; this trend was also verified by the validation cohort. Additionally, the survival model incorporating a composite of PLNN and TNM classification (C-index, 0.705, 0.734) performed better than the PLNN and AJCC TNM models. PLNN can serve as a powerful survival predictor for patients with HPSCC and is a surrogate supplement for cancer staging systems.

Keywords: hypopharyngeal squamous cell carcinoma, positive lymph nodes number, prognosis, prediction models, survival predictive values

Procedia PDF Downloads 154

Authors: Alan Wan

Abstract:

In recent years, the body of literature on frequentist model averaging in statistics has grown significantly. Most of this work focuses on models with different mean structures but leaves out the variance consideration. In this paper, we consider a regression model with multiplicative heteroscedasticity and develop a model averaging method that combines maximum likelihood estimators of unknown parameters in both the mean and variance functions of the model. Our weight choice criterion is based on a minimisation of a plug-in estimator of the model average estimator's squared prediction risk. We prove that the new estimator possesses an asymptotic optimality property. Our investigation of finite-sample performance by simulations demonstrates that the new estimator frequently exhibits very favourable properties compared to some existing heteroscedasticity-robust model average estimators. The model averaging method hedges against the selection of very bad models and serves as a remedy to variance function misspecification, which often discourages practitioners from modeling heteroscedasticity altogether. The proposed model average estimator is applied to the analysis of two real data sets.

Keywords: heteroscedasticity-robust, model averaging, multiplicative heteroscedasticity, plug-in, squared prediction risk

Procedia PDF Downloads 385

Authors: Andrea ČopíKová

Abstract:

Competency Based Management is a new approach to management, which solves organization’s challenges with complexity and with the aim to find and solve organization’s problems and learn how to avoid these in future. They teach the organizations to create, apart from the state of stability – that is temporary, vital organization, which is permanently able to utilize and profit from internal and external opportunities. The aim of this paper is to propose a process of competency model design, based on which a competency model for a financial department manager in a production company will be created. Competency models are very useful tool in many personnel processes in any organization. They are used for acquiring and selection of employees, designing training and development activities, employees’ evaluation, and they can be used as a guide for a career planning and as a tool for succession planning especially for managerial positions. When creating a competency model the method AHP (Analytic Hierarchy Process) and quantitative pair-wise comparison (Saaty’s method) will be used; these methods belong among the most used methods for the determination of weights, and it is used in the AHP procedure. The introduction part of the paper consists of the research results pertaining to the use of competency model in practice and then the issue of competency and competency models is explained. The application part describes in detail proposed methodology for the creation of competency models, based on which the competency model for the position of financial department manager in a foreign manufacturing company, will be created. In the conclusion of the paper, the final competency model will be shown for above mentioned position. The competency model divides selected competencies into three groups that are managerial, interpersonal and functional. The model describes in detail individual levels of competencies, their target value (required level) and the level of importance.

Keywords: analytic hierarchy process, competency, competency model, quantitative pairwise comparison

Procedia PDF Downloads 244

Authors: Ying Han, Yuanxiang Chen, Yongtao Huang, Jia Fu, Kaile Li, Shangjing Lin, Jianguo Yu

Abstract:

In long-haul and high-speed optical transmission system, the orthogonal frequency division multiplexing (OFDM) signal suffers various linear and non-linear impairments. In recent years, researchers have proposed compensation schemes for specific impairment, and the effects are remarkable. However, different impairment compensation algorithms have caused an increase in transmission delay. With the widespread application of deep neural networks (DNN) in communication, multi-impairment compensation based on DNN will be a promising scheme. In this paper, we propose and apply DNN to compensate multi-impairment of 16-QAM coherent optical OFDM signal, thereby improving the performance of the transmission system. The trained DNN models are applied in the offline digital signal processing (DSP) module of the transmission system. The models can optimize the constellation mapping signals at the transmitter and compensate multi-impairment of the OFDM decoded signal at the receiver. Furthermore, the models reduce the peak to average power ratio (PAPR) of the transmitted OFDM signal and the bit error rate (BER) of the received signal. We verify the effectiveness of the proposed scheme for 16-QAM Coherent Optical OFDM signal and demonstrate and analyze transmission performance in different transmission scenarios. The experimental results show that the PAPR and BER of the transmission system are significantly reduced after using the trained DNN. It shows that the DNN with specific loss function and network structure can optimize the transmitted signal and learn the channel feature and compensate for multi-impairment in fiber transmission effectively.

Keywords: coherent optical OFDM, deep neural network, multi-impairment compensation, optical transmission

Procedia PDF Downloads 143

Authors: Chen-Yin Kuo, Yung-Hsin Lee

Abstract:

Previous studies applying residual income valuation (RIV) model generally use panel data and single-equation model to forecast stock prices. Unlike these, this paper uses Taiwan longitudinal data to estimate multi-equation time-series models such as Vector Autoregressive (VAR), Vector Error Correction Model (VECM), and conduct out-of-sample forecasting. Further, this work assesses their forecasting performance by two instruments. In favor of extant research, the major finding shows that VECM outperforms other three models in forecasting for three stock sectors over entire horizons. It implies that an error correction term containing long-run information contributes to improve forecasting accuracy. Moreover, the pattern of composite shows that at longer horizon, VECM produces the greater reduction in errors, and performs substantially better than VAR.

Keywords: residual income valuation model, vector error correction model, out of sample forecasting, forecasting accuracy

Procedia PDF Downloads 316

Authors: Sourabh Jain, Parul Madan

Abstract:

Traffic noise pollution has become a challenging problem for all metro cities of India due to rapid urbanization, growing population and rising number of vehicles and transport development. In Delhi the prime source of noise pollution is vehicular traffic. In Delhi it is found that the ambient noise level (Leq) is exceeding the standard permissible value at all the locations. Noise barriers or enclosures are definitely useful in obtaining effective deduction of traffic noise disturbances in urbanized areas. US’s Federal Highway Administration Model (FHWA) and Calculation of Road Traffic Noise (CORTN) of UK are used to develop spread sheets for noise prediction. Spread sheets are also developed for evaluating effectiveness of existing boundary walls abutting houses in mitigating noise, redesigning them as noise barriers. Study was also carried out to examine the changes in noise level due to designed noise barrier by using both models FHWA and CORTN respectively. During the collection of various data it is found that receivers are located far away from road at Rithala and Moolchand sites and hence extra barrier height needed to meet prescribed limits was less as seen from calculations and most of the noise diminishes by propagation effect.On the basis of overall study and data analysis, it is concluded that FHWA and CORTN models under estimate noise levels. FHWA model predicted noise levels with an average percentage error of -7.33 and CORTN predicted with an average percentage error of -8.5. It was observed that at all sites noise levels at receivers were exceeding the standard limit of 55 dB. It was seen from calculations that existing walls are reducing noise levels. Average noise reduction due to walls at Rithala was 7.41 dB and at Panchsheel was 7.20 dB and lower amount of noise reduction was observed at Friend colony which was only 5.88. It was observed from analysis that Friends colony sites need much greater height of barrier. This was because of residential buildings abutting the road. At friends colony great amount of traffic was observed since it is national highway. At this site diminishing of noise due to propagation effect was very less.As FHWA and CORTN models were developed in excel programme, it eliminates laborious calculations of noise. There was no reflection correction in FHWA models as like in CORTN model.

Keywords: IFHWA, CORTN, Noise Sources, Noise Barriers

Procedia PDF Downloads 133

Authors: Qazi Salman Khalid, Muhammad Khalid, Shahid Maqsood

Abstract:

This paper presents a solution for optimizing the cycle time in an assembly line with human-robot collaboration and diverse operators. A genetic algorithm with tailored parameters is used to address the assembly line balancing problem in the automobile sector. A mathematical model is developed, depicting the problem. Currently, the firm runs on the largest candidate rule; however, it causes a lag in orders, which ultimately gets penalized. The results of the study show that the proposed GA is effective in providing efficient solutions and that the cycle time has significantly impacted productivity.

Keywords: line balancing, cycle time, genetic algorithm, productivity

Procedia PDF Downloads 137

Authors: Eva Nedeliakova, Juraj Camaj, Jaroslav Masek

Abstract:

This contribution is focused on the methodology for identifying levels of quality and improving quality through new logistics model in railway transport. It is oriented on the application of dynamic quality models, which represent an innovative method of evaluation quality services. Through this conception, time factor, expected, and perceived quality in each moment of the transportation process within logistics chain can be taken into account. Various models describe the improvement of the quality which emphases the time factor throughout the whole transportation logistics chain. Quality of services in railway transport can be determined by the existing level of service quality, by detecting the causes of dissatisfaction employees but also customers, to uncover strengths and weaknesses. This new logistics model is able to recognize critical processes in logistic chain. It includes service quality rating that must respect its specific properties, which are unrepeatability, impalpability, their use right at the time they are provided and particularly changeability, which is significant factor in the conditions of rail transport as well. These peculiarities influence the quality of service regarding the constantly increasing requirements and that result in new ways of finding progressive attitudes towards the service quality rating.

Keywords: logistics model, quality, railway transport

Procedia PDF Downloads 568

Authors: Chad Goldsworthy, B. Rajeswari Matam

Abstract:

The use of deep learning for species identification in camera trap images has revolutionised our ability to study, conserve and monitor species in a highly efficient and unobtrusive manner, with state-of-the-art models achieving accuracies surpassing the accuracy of manual human classification. The high imbalance of camera trap datasets, however, results in poor accuracies for minority (rare or endangered) species due to their relative insignificance to the overall model accuracy. This paper investigates the use of Focal Loss, in comparison to the traditional Cross Entropy Loss function, to improve the identification of minority species in the “255 Bird Species” dataset from Kaggle. The results show that, although Focal Loss slightly decreased the accuracy of the majority species, it was able to increase the F1-score by 0.06 and improve the identification of the bottom two, five and ten (minority) species by 37.5%, 15.7% and 10.8%, respectively, as well as resulting in an improved overall accuracy of 2.96%.

Keywords: convolutional neural networks, data imbalance, deep learning, focal loss, species classification, wildlife conservation

Procedia PDF Downloads 191

Authors: Maryam Zarabian, Hector Guzman, Pedro Pereira-Almao, Abraham Fapojuwo

Abstract:

Dry reforming of methane (DRM) has sparked significant industrial and scientific interest not only as a viable alternative for addressing the environmental concerns of two main contributors of the greenhouse effect, i.e., carbon dioxide (CO₂) and methane (CH₄), but also produces syngas, i.e., a mixture of hydrogen (H₂) and carbon monoxide (CO) utilized by a wide range of downstream processes as a feedstock for other chemical productions. In this study, we develop an AI-enable syngas production model to tackle the problem of achieving an equivalent H₂/CO ratio [1:1] with respect to the most efficient conversion. Firstly, the unsupervised density-based spatial clustering of applications with noise (DBSAN) algorithm removes outlier data points from the original experimental dataset. Then, random forest (RF) and deep neural network (DNN) models employ the error-free dataset to predict the DRM results. DNN models inherently would not be able to obtain accurate predictions without a huge dataset. To cope with this limitation, we employ reusing pre-trained layers’ approaches such as transfer learning and greedy layer-wise pretraining. Compared to the other deep models (i.e., pure deep model and transferred deep model), the greedy layer-wise pre-trained deep model provides the most accurate prediction as well as similar accuracy to the RF model with R² values 1.00, 0.999, 0.999, 0.999, 0.999, and 0.999 for the total outlet flow, H₂/CO ratio, H₂ yield, CO yield, CH₄ conversion, and CO₂ conversion outputs, respectively.

Keywords: artificial intelligence, dry reforming of methane, artificial neural network, deep learning, machine learning, transfer learning, greedy layer-wise pretraining

Procedia PDF Downloads 86

Authors: Josef Novák, Alena Kohoutková

Abstract:

The performance of concrete structures in fire depends on several factors which include, among others, the change in material properties due to the fire. Today, fiber reinforced concrete (FRC) belongs to materials which have been widely used for various structures and elements. While the knowledge and experience with FRC behavior under ambient temperature is well-known, the effect of elevated temperature on its behavior has to be deeply investigated. This paper deals with an experimental investigation and stress‑strain relations for hybrid fiber reinforced concrete (HFRC) which contains siliceous aggregates, polypropylene and steel fibers. The main objective of the experimental investigation is to enhance a database of mechanical properties of concrete composites with addition of fibers subject to elevated temperature as well as to validate existing stress-strain relations for HFRC. Within the investigation, a unique heat transport test, compressive test and splitting tensile test were performed on 150 mm cubes heated up to 200, 400, and 600 °C with the aim to determine a time period for uniform heat distribution in test specimens and the mechanical properties of the investigated concrete composite, respectively. Both findings obtained from the presented experimental test as well as experimental data collected from scientific papers so far served for validating the computational accuracy of investigated stress-strain relations for HFRC which have been developed during last few years. Owing to the presence of steel and polypropylene fibers, HFRC becomes a unique material whose structural performance differs from conventional plain concrete when exposed to elevated temperature. Polypropylene fibers in HFRC lower the risk of concrete spalling as the fibers burn out shortly with increasing temperature due to low ignition point and as a consequence pore pressure decreases. On the contrary, the increase in the concrete porosity might affect the mechanical properties of the material. To validate this thought requires enhancing the existing result database which is very limited and does not contain enough data. As a result of the poor database, only few stress-strain relations have been developed so far to describe the structural performance of HFRC at elevated temperature. Moreover, many of them are inconsistent and need to be refined. Most of them also do not take into account the effect of both a fiber type and fiber content. Such approach might be vague especially when high amount of polypropylene fibers are used. Therefore, the existing relations should be validated in detail based on other experimental results.

Keywords: elevated temperature, fiber reinforced concrete, mechanical properties, stress strain relation

Procedia PDF Downloads 339

Authors: Wenhao Li, Shijun Guo

Abstract:

Carbon fibre reinforced composites are progressively replacing metal structures in modern civil aircraft. This is because composite materials have large potential of weight saving compared with metal. However, the achievement to date of weight saving in composite structure is far less than the theoretical potential due to many uncertainties in structural integrity and safety concern. Unlike the conventional metallic structure, composite components are bonded together along the joints where structural integrity is a major concern. To ensure the safety, metal fasteners are used to reinforce the composite bonded joints. One of the solutions for a significant weight saving of composite structure is to develop an effective technology of on-board Structural Health Monitoring (SHM) System. By monitoring the real-life stress status of composite structures during service, the safety margin set in the structure design can be reduced with confidence. It provides a means of safeguard to minimize the need for programmed inspections and allow for maintenance to be need-driven, rather than usage-driven. The aim of this paper is to develop smart composite joint. The key technology is a multifunctional thermoplastic composite fastener (MTCF). The MTCF will replace some of the existing metallic fasteners in the most concerned locations distributed over the aircraft composite structures to reinforce the joints and form an on-board SHM network system. Each of the MTCFs will work as a unit of the AU and AE technology. The proposed MTCF technology has been patented and developed by Prof. Guo in Cranfield University, UK in the past a few years. The manufactured MTCF has been successfully employed in the composite SLJ (Single-Lap Joint). In terms of the structure integrity, the hybrid SLJ reinforced by MTCF achieves 19.1% improvement in the ultimate failure strength in comparison to the bonded SLJ. By increasing the diameter or rearranging the lay-up sequence of MTCF, the hybrid SLJ reinforced by MTCF is able to achieve the equivalent ultimate strength as that reinforced by titanium fastener. The predicted ultimate strength in simulation is in good agreement with the test results. In terms of the structural health monitoring, a signal from the MTCF was measured well before the load of mechanical failure. This signal provides a warning of initial crack in the joint which could not be detected by the strain gauge until the final failure.

Keywords: composite single-lap joint, crack propagation, multifunctional composite fastener, structural health monitoring

Procedia PDF Downloads 163

Authors: W. Sriwattanapongse, S. Prasitwattanaseree, S. Wongtrangan

Abstract:

The transportation accidents mortality is a major problem that leads to loss of human lives, and economic. The objective was to identify patterns of statistical modeling for estimating mortality rates due to transportation accidents in Thailand by using data from 2000 to 2009. The data was taken from the death certificate, vital registration database. The number of deaths and mortality rates were computed classifying by gender, age, year and region. There were 114,790 cases of transportation accidents deaths. The highest average age-specific transport accident mortality rate is 3.11 per 100,000 per year in males, Southern region and the lowest average age-specific transport accident mortality rate is 1.79 per 100,000 per year in females, North-East region. Linear, poisson and negative binomial models were chosen for fitting statistical model. Among the models fitted, the best was chosen based on the analysis of deviance and AIC. The negative binomial model was clearly appropriate fitted.

Keywords: transportation accidents, mortality, modeling, analysis of deviance

Procedia PDF Downloads 244

Authors: M. Ezzat, Y. Zaghloul, T. Sorour, A. Hefny, M. Eid

Abstract:

Ultimate capacity of large diameter bored piles is usually determined from pile loading tests as recommended by several international codes and foundation design standards. However, loading of this type of piles till achieving apparent failure is practically seldom. In this paper, numerical analyses are carried out to simulate load test of a large diameter bored pile performed at the location of Alzey highway bridge project (Germany). Test results of pile load settlement relationship till failure as well as results of the base and shaft resistances are available. Apparent failure was indicated in this test by the significant increase of the induced settlement during the last load increment applied on the pile head. Measurements of this pile load test are used to assess the quality of the numerical models investigated. Three different material soil models are implemented in the analyses: Mohr coulomb (MC), Soft soil (SS), and Modified Mohr coulomb (MMC). Very good agreement is obtained between the field measured settlement and the calculated settlement using the MMC model. Results of analysis showed also that the MMC constitutive model is superior to MC, and SS models in predicting the ultimate base and shaft resistances of the large diameter bored pile. After calibrating the numerical model, behavior of large diameter bored piles under axial loads is discussed and the formation of the plastic zone around the pile is explored. Results obtained showed that the plastic zone below the base of the pile at failure extended laterally to about four times the pile diameter and vertically to about three times the pile diameter.

Keywords: ultimate capacity, large diameter bored piles, plastic zone, failure, pile load test

Procedia PDF Downloads 143

Authors: Nadjib Melkemi, Salah Belaidi

Abstract:

Quantitative Structure-Activity Relationship (QSAR) studies have been performed on nineteen molecules of 1,2-dithiole-3-thione analogues. The compounds used are the potent inducers of enzymes involved in the maintenance of reduced glutathione pools as well as phase-2 enzymes important to electrophile detoxication. A multiple linear regression (MLR) procedure was used to design the relationships between molecular descriptor and detoxication properties of the 1,2-dithiole-3-thione derivatives. The predictivity of the model was estimated by cross-validation with the leave-one-out method. Our results suggest a QSAR model based of the following descriptors: qS2, qC3, qC5, qS6, DM, Pol, log P, MV, SAG, HE and EHOMO for the specific activity of quinone reductase; qS1, qS2, qC3, qC4, qC5, qS6, DM, Pol, logP, MV, SAG, HE and EHOMO for the production of growth hormone. To confirm the predictive power of the models, an external set of molecules was used. High correlation between experimental and predicted activity values was observed, indicating the validation and the good quality of the derived QSAR models.

Keywords: QSAR, quinone reductase activity, production of growth hormone, MLR

Procedia PDF Downloads 350