Search results for: converting models

Authors: William Albeiro Alvarez, Gloria Patricia Jaramillo, Ivan Reinaldo Sarmiento

Abstract:

Understanding the behavior of individuals and the different human factors that influence the choice when faced with a complex system such as transportation is one of the most complicated aspects of measuring in the components that constitute the modeling of route choice due to that various behaviors and driving mode directly or indirectly affect the choice. During the last two decades, with the development of information and communications technologies, new data collection techniques have emerged such as GPS, geolocation with mobile phones, apps for choosing the route between origin and destination, individual service transport applications among others, where an interest has been generated to improve discrete choice models when considering the incorporation of these developments as well as psychological factors that affect decision making. This paper implements a discrete choice model that proposes and estimates a hybrid model that integrates route choice models and latent variables based on the observation on the route of a sample of public taxi drivers from the city of Medellín, Colombia in relation to its behavior, personality, socioeconomic characteristics, and driving mode. The set of choice options includes the routes generated by the individual service transport applications versus the driver's choice. The hybrid model consists of measurement equations that relate latent variables with measurement indicators and utilities with choice indicators along with structural equations that link the observable characteristics of drivers with latent variables and explanatory variables with utilities.

Keywords: behavior choice model, human factors, hybrid model, real time data

Procedia PDF Downloads 151

Authors: Min Jindong, Wang Mingxia

Abstract:

With the widespread use of artificial intelligence in life, computer vision, especially deep convolutional neural network models, has developed rapidly. With the increase of the complexity of the real visual target detection task and the improvement of the recognition accuracy, the target detection network model is also very large. The huge deep neural network model is not conducive to deployment on edge devices with limited resources, and the timeliness of network model inference is poor. In this paper, knowledge distillation is used to compress the huge and complex deep neural network model, and the knowledge contained in the complex network model is comprehensively transferred to another lightweight network model. Different from traditional knowledge distillation methods, we propose a novel knowledge distillation that incorporates multi-faceted features, called M-KD. In this paper, when training and optimizing the deep neural network model for target detection, the knowledge of the soft target output of the teacher network in knowledge distillation, the relationship between the layers of the teacher network and the feature attention map of the hidden layer of the teacher network are transferred to the student network as all knowledge. in the model. At the same time, we also introduce an intermediate transition layer, that is, an intermediate guidance layer, between the teacher network and the student network to make up for the huge difference between the teacher network and the student network. Finally, this paper adds an exploration module to the traditional knowledge distillation teacher-student network model. The student network model not only inherits the knowledge of the teacher network but also explores some new knowledge and characteristics. Comprehensive experiments in this paper using different distillation parameter configurations across multiple datasets and convolutional neural network models demonstrate that our proposed new network model achieves substantial improvements in speed and accuracy performance.

Keywords: object detection, knowledge distillation, convolutional network, model compression

Procedia PDF Downloads 276

Authors: M. Govindarajan

Abstract:

One of the major developments in machine learning in the past decade is the ensemble method, which finds highly accurate classifier by combining many moderately accurate component classifiers. In this research work, new ensemble classification methods are proposed with homogeneous ensemble classifier using bagging and heterogeneous ensemble classifier using arcing and their performances are analyzed in terms of accuracy. A Classifier ensemble is designed using Radial Basis Function (RBF) and Support Vector Machine (SVM) as base classifiers. The feasibility and the benefits of the proposed approaches are demonstrated by the means of standard datasets of intrusion detection. The main originality of the proposed approach is based on three main parts: preprocessing phase, classification phase, and combining phase. A wide range of comparative experiments is conducted for standard datasets of intrusion detection. The performance of the proposed homogeneous and heterogeneous ensemble classifiers are compared to the performance of other standard homogeneous and heterogeneous ensemble methods. The standard homogeneous ensemble methods include Error correcting output codes, Dagging and heterogeneous ensemble methods include majority voting, stacking. The proposed ensemble methods provide significant improvement of accuracy compared to individual classifiers and the proposed bagged RBF and SVM performs significantly better than ECOC and Dagging and the proposed hybrid RBF-SVM performs significantly better than voting and stacking. Also heterogeneous models exhibit better results than homogeneous models for standard datasets of intrusion detection. 

Keywords: data mining, ensemble, radial basis function, support vector machine, accuracy

Procedia PDF Downloads 246

Authors: Mario Perez-Won, Roberto Lemus-Mondaca, Fernanda Marin, Constanza Olivares

Abstract:

This study presents the simultaneous application of high hydrostatic pressure (HHP) and osmotic dehydration of jumbo squid (Dosidicus gigas) slice. Diffusion coefficients for both components water and solids were improved by the process pressure, being influenced by pressure level. The working conditions were different pressures such as 100, 250, 400 MPa and pressure atmospheric (0.1 MPa) for time intervals from 30 to 300 seconds and a 15% NaCl concentration. The mathematical expressions used for mass transfer simulations both water and salt were those corresponding to Newton, Henderson and Pabis, Page and Weibull models, where the Weibull and Henderson-Pabis models presented the best fitted to the water and salt experimental data, respectively. The values for water diffusivity coefficients varied from 1.62 to 8.10x10⁻⁹ m²/s whereas that for salt varied among 14.18 to 36.07x10⁻⁹ m²/s for selected conditions. Finally, as to quality parameters studied under the range of experimental conditions studied, the treatment at 250 MPa yielded on the samples a minimum hardness, whereas springiness, cohesiveness and chewiness at 100, 250 and 400 MPa treatments presented statistical differences regarding to unpressurized samples. The colour parameters L* (lightness) increased, however, but b* (yellowish) and a* (reddish) parameters decreased when increasing pressure level. This way, samples presented a brighter aspect and a mildly cooked appearance. The results presented in this study can support the enormous potential of hydrostatic pressure application as a technique important for compounds impregnation under high pressure.

Keywords: colour, diffusivity, high pressure, jumbo squid, modelling, texture

Procedia PDF Downloads 343

Authors: Majid Saaly, Shahriar Tavousi Tafreshi, Mehdi Nazari Afshar

Abstract:

Most reinforced concrete structures are designed only under heavy loads have large transverse reinforcement spacing values, and therefore suffer severe failure after intense ground movements. The main goal of this paper is to compare the shear- and axial failure of concrete bending frames available in Tehran using incremental dynamic analysis under near- and far-field records. For this purpose, IDA analyses of 5, 10, and 15-story concrete structures were done under seven far-fault records and five near-faults records. The results show that in two-dimensional models of short-rise, mid-rise and high-rise reinforced concrete frames located on Type-3 soil, increasing the distance of the transverse reinforcement can increase the maximum inter-story drift ratio values up to 37%. According to the existing results on 5, 10, and 15-story reinforced concrete models located on Type-3 soil, records with characteristics such as fling-step and directivity create maximum drift values between floors more than far-fault earthquakes. The results indicated that in the case of seismic excitation modes under earthquake encompassing directivity or fling-step, the probability values of failure and failure possibility increasing rate values are much smaller than the corresponding values of far-fault earthquakes. However, in near-fault frame records, the probability of exceedance occurs at lower seismic intensities compared to far-fault records.

Keywords: IDA, failure curve, directivity, maximum floor drift, fling step, evaluating probable bending of frames, near-field and far-field earthquake records

Procedia PDF Downloads 103

Authors: Ahmed Mantawy, James C. Anderson

Abstract:

Numerous cases of earthquakes (main-shocks) that were followed by aftershocks have been recorded in California. In 1992 a pair of strong earthquakes occurred within three hours of each other in Southern California. The first shock occurred near the community of Landers and was assigned a magnitude of 7.3 then the second shock occurred near the city of Big Bear about 20 miles west of the initial shock and was assigned a magnitude of 6.2. In the same year, a series of three earthquakes occurred over two days in the Cape-Mendocino area of Northern California. The main-shock was assigned a magnitude of 7.0 while the second and the third shocks were both assigned a value of 6.6. This paper investigates the effect of a main-shock accompanied with aftershocks of significant intensity on reinforced concrete (RC) frame buildings to indicate nonlinear behavior using PERFORM-3D software. A 6-story building in San Bruno and a 20-story building in North Hollywood were selected for the study as both of them have RC moment resisting frame systems. The buildings are also instrumented at multiple floor levels as a part of the California Strong Motion Instrumentation Program (CSMIP). Both buildings have recorded responses during past events such as Loma-Prieta and Northridge earthquakes which were used in verifying the response parameters of the numerical models in PERFORM-3D. The verification of the numerical models shows good agreement between the calculated and the recorded response values. Then, different scenarios of a main-shock followed by a series of aftershocks from real cases in California were applied to the building models in order to investigate the structural behavior of the moment-resisting frame system. The behavior was evaluated in terms of the lateral floor displacements, the ductility demands, and the inelastic behavior at critical locations. The analysis results showed that permanent displacements may have happened due to the plastic deformation during the main-shock that can lead to higher displacements during after-shocks. Also, the inelastic response at plastic hinges during the main-shock can change the hysteretic behavior during the aftershocks. Higher ductility demands can also occur when buildings are subjected to trains of ground motions compared to the case of individual ground motions. A general conclusion is that the occurrence of aftershocks following an earthquake can lead to increased damage within the elements of an RC frame buildings. Current code provisions for seismic design do not consider the probability of significant aftershocks when designing a new building in zones of high seismic activity.

Keywords: reinforced concrete, existing buildings, aftershocks, damage accumulation

Procedia PDF Downloads 279

Authors: Lovesh Goyal, Pushkar Dave, Prajyot Jadhav, GonnaYaswanth, Sakshi Giri, Sahil Dharme, Rushika Joshi, Rishabh Verma, Shital Chiddarwar

Abstract:

With the rising need for human-robot interaction, researchers have proposed and tested multiple models with varying degrees of success. A few of these models performed on aerial platforms are commonly known as Tethered Aerial Systems. These aerial vehicles may be powered continuously by a tether cable, which addresses the predicament of the short battery life of quadcopters. This system finds applications to minimize humanitarian efforts for industrial, medical, agricultural, and service uses. However, a significant challenge in employing such systems is that it necessities attaining smooth and secure robot-human interaction while ensuring that the forces from the tether remain within the standard comfortable range for the humans. To tackle this problem, a hybrid control method that could switch between two control techniques: constant control input and the steady-state solution, is implemented. The constant control approach is implemented when a person is far from the target location, and error is thought to be eventually constant. The controller switches to the steady-state approach when the person reaches within a specific range of the goal position. Both strategies take into account human velocity feedback. This hybrid technique enhances the outcomes by assisting the person to reach the desired location while decreasing the human's unwanted disturbance throughout the process, thereby keeping the interaction between the robot and the subject smooth.

Keywords: unmanned aerial vehicle, tethered system, physical human-robot interaction, hybrid control

Procedia PDF Downloads 95

Authors: Jiří F. Urbánek, Alena Oulehlová, Hana Malachová, Jiří J. Urbánek Jr.

Abstract:

Crisis situations investigation and modelling are introduced and made within the complex system of energetic critical infrastructure, operating on peril´s environments. Every crisis situations and perils has an origin in the emergency/ crisis event occurrence and they need critical/ crisis interfaces assessment. Here, the emergency events can be expected - then crisis scenarios can be pre-prepared by pertinent organizational crisis management authorities towards their coping; or it may be unexpected - without pre-prepared scenario of event. But the both need operational coping by means of crisis management as well. The operation, forms, characteristics, behaviour and utilization of crisis management have various qualities, depending on real critical infrastructure organization perils, and prevention training processes. An aim is always - better security and continuity of the organization, which successful obtainment needs to find and investigate critical/ crisis zones and functions in critical infrastructure organization models, operating in pertinent perils environment. Our DYVELOP (Dynamic Vector Logistics of Processes) method is disposables for it. Here, it is necessary to derive and create identification algorithm of critical/ crisis interfaces. The locations of critical/ crisis interfaces are the flags of crisis situation in organization of critical infrastructure models. Then, the model of crisis situation will be displayed at real organization of Czech energetic crisis infrastructure subject in real peril environment. These efficient measures are necessary for the infrastructure protection. They will be derived for peril mitigation, crisis situation coping and for environmentally friendly organization survival, continuity and its sustainable development advanced possibilities.

Keywords: algorithms, energetic infrastructure complex system, modelling, peril´s environment

Procedia PDF Downloads 402

Authors: Bitewulign Mekonnen

Abstract:

Context: This scientific paper focuses on the use of near-infrared (NIR) spectroscopy to determine glucose concentration in aqueous solutions accurately and rapidly. The study compares six different machine learning methods for predicting glucose concentration and also explores the development of a deep learning model for classifying NIR spectra. The objective is to optimize the detection model and improve the accuracy of glucose prediction. This research is important because it provides a comprehensive analysis of various machine-learning techniques for estimating aqueous glucose concentrations. Research Aim: The aim of this study is to compare and evaluate different machine-learning methods for predicting glucose concentration from NIR spectra. Additionally, the study aims to develop and assess a deep-learning model for classifying NIR spectra. Methodology: The research methodology involves the use of machine learning and deep learning techniques. Six machine learning regression models, including support vector machine regression, partial least squares regression, extra tree regression, random forest regression, extreme gradient boosting, and principal component analysis-neural network, are employed to predict glucose concentration. The NIR spectra data is randomly divided into train and test sets, and the process is repeated ten times to increase generalization ability. In addition, a convolutional neural network is developed for classifying NIR spectra. Findings: The study reveals that the SVMR, ETR, and PCA-NN models exhibit excellent performance in predicting glucose concentration, with correlation coefficients (R) > 0.99 and determination coefficients (R²)> 0.985. The deep learning model achieves high macro-averaging scores for precision, recall, and F1-measure. These findings demonstrate the effectiveness of machine learning and deep learning methods in optimizing the detection model and improving glucose prediction accuracy. Theoretical Importance: This research contributes to the field by providing a comprehensive analysis of various machine-learning techniques for estimating glucose concentrations from NIR spectra. It also explores the use of deep learning for the classification of indistinguishable NIR spectra. The findings highlight the potential of machine learning and deep learning in enhancing the prediction accuracy of glucose-relevant features. Data Collection and Analysis Procedures: The NIR spectra and corresponding references for glucose concentration are measured in increments of 20 mg/dl. The data is randomly divided into train and test sets, and the models are evaluated using regression analysis and classification metrics. The performance of each model is assessed based on correlation coefficients, determination coefficients, precision, recall, and F1-measure. Question Addressed: The study addresses the question of whether machine learning and deep learning methods can optimize the detection model and improve the accuracy of glucose prediction from NIR spectra. Conclusion: The research demonstrates that machine learning and deep learning methods can effectively predict glucose concentration from NIR spectra. The SVMR, ETR, and PCA-NN models exhibit superior performance, while the deep learning model achieves high classification scores. These findings suggest that machine learning and deep learning techniques can be used to improve the prediction accuracy of glucose-relevant features. Further research is needed to explore their clinical utility in analyzing complex matrices, such as blood glucose levels.

Keywords: machine learning, signal processing, near-infrared spectroscopy, support vector machine, neural network

Procedia PDF Downloads 92

Authors: Mohammad Ghiasvand, Babak Khorsandi, Morteza Kolahdoozan

Abstract:

Under the influence of waves, oil in the sea is subject to vertical scattering in the water column. Scientists' knowledge of how oil is dispersed in the water column is one of the lowest levels of knowledge among other processes affecting oil in the marine environment, which highlights the need for research and study in this field. Therefore, this study investigates the distribution of oil in the water column in a turbulent environment with zero velocity characteristics. Lack of laboratory results to analyze the distribution of petroleum pollutants in deep water for information Phenomenon physics on the one hand and using them to calibrate numerical models on the other hand led to the development of laboratory models in research. According to the aim of the present study, which is to investigate the distribution of oil in homogeneous and isotropic turbulence caused by the oscillating Grid, after reaching the ideal conditions, the crude oil flow was poured onto the water surface and oil was distributed in deep water due to turbulence was investigated. In this study, all experimental processes have been implemented and used for the first time in Iran, and the study of oil diffusion in the water column was considered one of the key aspects of pollutant diffusion in the oscillating Grid environment. Finally, the required oscillation velocities were taken at depths of 10, 15, 20, and 25 cm from the water surface and used in the analysis of oil diffusion due to turbulence parameters. The results showed that with the characteristics of the present system in two static modes and network motion with a frequency of 0.8 Hz, the results of oil diffusion in the four mentioned depths at a frequency of 0.8 Hz compared to the static mode from top to bottom at 26.18, 57 31.5, 37.5 and 50% more. Also, after 2.5 minutes of the oil spill at a frequency of 0.8 Hz, oil distribution at the mentioned depths increased by 49, 61.5, 85, and 146.1%, respectively, compared to the base (static) state.

Keywords: homogeneous and isotropic turbulence, oil distribution, oscillating grid, oil spill

Procedia PDF Downloads 73

Authors: Yue Zhang, Bo Yang, Gang Xiong, Mohamed Elchalakanic, Shidong Nie

Abstract:

This study aims to investigate the lateral torsional buckling of I-shaped cross-section beams fabricated from Q460GJ structural steel plates. Both experimental and numerical simulation results are presented in this paper. A total of eight specimens were tested under a three-point bending, and the corresponding numerical models were established to conduct parametric studies. The effects of some key parameters such as the non-dimensional member slenderness and the height-to-width ratio, were investigated based on the verified numerical models. Also, the results obtained from the parametric studies were compared with the predictions calculated by different design codes including the Chinese design code (GB50017-2003, 2003), the new draft version of Chinese design code (GB50017-201X, 2012), Eurocode 3 (EC3, 2005) and the North America design code (ANSI/AISC360-10, 2010). These comparisons indicated that the sectional height-to-width ratio does not play an important role to influence the overall stability load-carrying capacity of Q460GJ structural steel beams with welded I-shaped cross-sections. It was also found that the design methods in GB50017-2003 and ANSI/AISC360-10 overestimate the overall stability and load-carrying capacity of Q460GJ welded I-shaped cross-section beams.

Keywords: experimental study, finite element analysis, global stability, lateral torsional buckling, Q460GJ structural steel

Procedia PDF Downloads 325

Authors: Piotr Dukalski, Bartlomiej Bedkowski, Tomasz Jarek, Tomasz Wolnik

Abstract:

The article is concerned with the design of an electric in wheel hub motor installed in an electric car with two-wheel drive. It presents the construction of the motor on the 3D cross-section model. Work simulation of the motor (applicated to Fiat Panda car) and selected driving parameters such as driving on the road with a slope of 20%, driving at maximum speed, maximum acceleration of the car from 0 to 100 km/h are considered by the authors in the article. The demand for the drive power taking into account the resistance to movement was determined for selected driving conditions. The parameters of the motor operation and the power losses in its individual elements, calculated using the FEM 2D method, are presented for the selected car driving parameters. The calculated power losses are used in 3D models for thermal calculations using the CFD method. Detailed construction of thermal models with materials data, boundary conditions and losses calculated using the FEM 2D method are presented in the article. The article presents and describes calculated temperature distributions in individual motor components such as winding, permanent magnets, magnetic core, body, cooling system components. Generated losses in individual motor components and their impact on the limitation of its operating parameters are described by authors. Attention is paid to the losses generated in permanent magnets, which are a source of heat as the removal of which from inside the motor is difficult. Presented results of calculations show how individual motor power losses, generated in different load conditions while driving, affect its thermal state.

Keywords: electric car, electric drive, electric motor, thermal calculations, wheel hub motor

Procedia PDF Downloads 172

Authors: Tanit Daniel Jodar Vecina, Adriane Prisco Petry

Abstract:

The shape of the wind velocity profile changes according to local features of terrain shape and roughness, which are parameters responsible for defining the Atmospheric Boundary Layer (ABL) profile. Air flow characteristics over and around landforms, such as hills, are of considerable importance for applications related to Wind Farm and Turbine Engineering. The air flow is accelerated on top of hills, which can represent a decisive factor for Wind Turbine placement choices. The present work focuses on the study of ABL behavior as a function of slope and surface roughness of hill-shaped landforms, using the Computational Fluid Dynamics (CFD) to build wind velocity and turbulent intensity profiles. Reynolds-Averaged Navier-Stokes (RANS) equations are closed using the SST k-ω turbulence model; numerical results are compared to experimental data measured in wind tunnel over scale models of the hills under consideration. Eight hill models with slopes varying from 25° to 68° were tested for two types of terrain categories in 2D and 3D, and two analytical codes are used to represent the inlet velocity profiles. Numerical results for the velocity profiles show differences under 4% when compared to their respective experimental data. Turbulent intensity profiles show maximum differences around 7% when compared to experimental data; this can be explained by not being possible to insert inlet turbulent intensity profiles in the simulations. Alternatively, constant values based on the averages of the turbulent intensity at the wind tunnel inlet were used.

Keywords: Atmospheric Boundary Layer, Computational Fluid Dynamic (CFD), Numerical Modeling, Wind Tunnel

Procedia PDF Downloads 379

Authors: Roman Graf, Sergiu Gordea, Heather M. Ryan

Abstract:

This paper presents an approach for the easy creation of an institutional risk profile for endangerment analysis of file formats. The main contribution of this work is the employment of data mining techniques to support risk factors set up with just the most important values that are important for a particular organisation. Subsequently, the risk profile employs fuzzy models and associated configurations for the file format metadata aggregator to support digital preservation experts with a semi-automatic estimation of endangerment level for file formats. Our goal is to make use of a domain expert knowledge base aggregated from a digital preservation survey in order to detect preservation risks for a particular institution. Another contribution is support for visualisation and analysis of risk factors for a requried dimension. The proposed methods improve the visibility of risk factor information and the quality of a digital preservation process. The presented approach is meant to facilitate decision making for the preservation of digital content in libraries and archives using domain expert knowledge and automatically aggregated file format metadata from linked open data sources. To facilitate decision-making, the aggregated information about the risk factors is presented as a multidimensional vector. The goal is to visualise particular dimensions of this vector for analysis by an expert. The sample risk profile calculation and the visualisation of some risk factor dimensions is presented in the evaluation section.

Keywords: digital information management, file format, endangerment analysis, fuzzy models

Procedia PDF Downloads 402

Authors: Pauline Bredy, Yves Schuurman, David Farrusseng

Abstract:

To fulfill climate objectives, the production of synthetic e-fuels using CO₂ as a raw material appears as part of the solution. In particular, Power-to-Liquid (PtL) concept combines CO₂ with hydrogen supplied from water electrolysis, powered by renewable sources, which is currently gaining interest as it allows the production of sustainable fossil-free liquid fuels. The proposed process discussed here is an upgrading of the well-known Fischer-Tropsch synthesis. The concept deals with two cascade reactions in one pot, with first the conversion of CO₂ into CO via the reverse water gas shift (RWGS) reaction, which is then followed by the Fischer-Tropsch Synthesis (FTS). Instead of using a Fe-based catalyst, which can carry out both reactions, we have chosen the strategy to decouple the two functions (RWGS and FT) on two different catalysts within the same reactor. The FTS shall shift the equilibrium of the RWGS reaction (which alone would be limited to 15-20% of conversion at 250°C) by converting the CO into hydrocarbons. This strategy shall enable optimization of the catalyst pair and thus lower the temperature of the reaction thanks to the equilibrium shift to gain selectivity in the liquid fraction. The challenge lies in maximizing the activity of the RWGS catalyst but also in the ability of the FT catalyst to be highly selective. Methane production is the main concern as the energetic barrier of CH₄ formation is generally lower than that of the RWGS reaction, so the goal will be to minimize methane selectivity. Here we report the study of different combinations of copper-based RWGS catalysts with different cobalt-based FTS catalysts. We investigated their behaviors under mild process conditions by the use of high-throughput experimentation. Our results show that at 250°C and 20 bars, Cobalt catalysts mainly act as methanation catalysts. Indeed, CH₄ selectivity never drops under 80% despite the addition of various protomers (Nb, K, Pt, Cu) on the catalyst and its coupling with active RWGS catalysts. However, we show that the activity of the RWGS catalyst has an impact and can lead to longer hydrocarbons chains selectivities (C₂⁺) of about 10%. We studied the influence of the reduction temperature on the activity and selectivity of the tandem catalyst system. Similar selectivity and conversion were obtained at reduction temperatures between 250-400°C. This leads to the question of the active phase of the cobalt catalysts, which is currently investigated by magnetic measurements and DRIFTS. Thus, in coupling it with a more selective FT catalyst, better results are expected. This was achieved using a cobalt/iron FTS catalyst. The CH₄ selectivity dropped to 62% at 265°C, 20 bars, and a GHSV of 2500ml/h/gcat. We propose that the conditions used for the cobalt catalysts could have generated this methanation because these catalysts are known to have their best performance around 210°C in classical FTS, whereas the iron catalysts are more flexible but are also known to have an RWGS activity.

Keywords: cobalt-copper catalytic systems, CO₂-hydrogenation, Fischer-Tropsch synthesis, hydrocarbons, low-temperature process

Procedia PDF Downloads 55

Authors: Faisal Muhibuddin, Ani Dijah Rahajoe

Abstract:

This in-depth exploration, through a systematic literature review, scrutinizes the nuanced role of data mining in the classification of criminal activities. The research focuses on investigating various methodological aspects and recent developments in leveraging data mining techniques to enhance the effectiveness and precision of crime categorization. Commencing with an exposition of the foundational concepts of crime classification and its evolutionary dynamics, this study details the paradigm shift from conventional methods towards approaches supported by data mining, addressing the challenges and complexities inherent in the modern crime landscape. Specifically, the research delves into various data mining techniques, including K-means clustering, Naïve Bayes, K-nearest neighbour, and clustering methods. A comprehensive review of the strengths and limitations of each technique provides insights into their respective contributions to improving crime classification models. The integration of diverse data sources takes centre stage in this research. A detailed analysis explores how the amalgamation of structured data (such as criminal records) and unstructured data (such as social media) can offer a holistic understanding of crime, enriching classification models with more profound insights. Furthermore, the study explores the temporal implications in crime classification, emphasizing the significance of considering temporal factors to comprehend long-term trends and seasonality. The availability of real-time data is also elucidated as a crucial element in enhancing responsiveness and accuracy in crime classification.

Keywords: data mining, classification algorithm, naïve bayes, k-means clustering, k-nearest neigbhor, crime, data analysis, sistematic literature review

Procedia PDF Downloads 62

Authors: Soumyadip Banerjee, Tanmoy Maity

Abstract:

The recovery of power from waste air generated by mine ventilation fans presents a promising avenue for enhancing energy efficiency in mining operations. This abstract explores the feasibility and benefits of utilizing turbine generators to capture the kinetic energy present in waste air and convert it into electrical power. By integrating turbine generator systems into mine ventilation infrastructures, the potential to harness and utilize the previously untapped energy within the waste air stream is realized. This study examines the principles underlying turbine generator technology and its application within the context of mine ventilation systems. The process involves directing waste air from ventilation fans through specially designed turbines, where the kinetic energy of the moving air is converted into rotational motion. This mechanical energy is then transferred to connected generators, which convert it into electrical power. The recovered electricity can be employed for various on-site applications, including powering mining equipment, lighting, and control systems. The benefits of power recovery from waste air using turbine generators are manifold. Improved energy efficiency within the mining environment results in reduced dependence on external power sources and associated cost savings. Additionally, this approach contributes to environmental sustainability by utilizing a previously wasted resource for power generation. Resource conservation is further enhanced, aligning with modern principles of sustainable mining practices. However, successful implementation requires careful consideration of factors such as waste air characteristics, turbine design, generator efficiency, and integration into existing mine infrastructure. Maintenance and monitoring protocols are necessary to ensure consistent performance and longevity of the turbine generator systems. While there is an initial investment associated with equipment procurement, installation, and integration, the long-term benefits of reduced energy costs and environmental impact make this approach economically viable. In conclusion, the recovery of power from waste air from mine ventilation fans using turbine generators offers a tangible solution to enhance energy efficiency and sustainability within mining operations. By capturing and converting the kinetic energy of waste air into usable electrical power, mines can optimize resource utilization, reduce operational costs, and contribute to a greener future for the mining industry.

Keywords: waste to energy, wind power generation, exhaust air, power recovery

Procedia PDF Downloads 32

Authors: Min Chen, Rubi Gupta

Abstract:

Understanding and predicting stock market movements is a challenging problem. It is believed stock markets are partially driven by public sentiments, which leads to numerous research efforts to predict stock market trend using public sentiments expressed on social media such as Twitter but with limited success. Recently a microblogging website StockTwits is becoming increasingly popular for users to share their discussions and sentiments about stocks and financial market. In this project, we analyze the text content of StockTwits tweets and extract financial sentiment using text featurization and machine learning algorithms. StockTwits tweets are first pre-processed using techniques including stopword removal, special character removal, and case normalization to remove noise. Features are extracted from these preprocessed tweets through text featurization process using bags of words, N-gram models, TF-IDF (term frequency-inverse document frequency), and latent semantic analysis. Machine learning models are then trained to classify the tweets' sentiment as positive (bullish) or negative (bearish). The correlation between the aggregated daily sentiment and daily stock price movement is then investigated using Pearson’s correlation coefficient. Finally, the sentiment information is applied together with time series stock data to predict stock price movement. The experiments on five companies (Apple, Amazon, General Electric, Microsoft, and Target) in a duration of nine months demonstrate the effectiveness of our study in improving the prediction accuracy.

Keywords: machine learning, sentiment analysis, stock price prediction, tweet processing

Procedia PDF Downloads 156

Authors: Mayowa A. Sofowora, Seraphin D. Eyono Obono

Abstract:

The use of information and communication technologies such as computers, mobile phones and the internet is becoming prevalent in today’s world; and it is facilitating access to a vast amount of data, services, and applications for the improvement of people’s lives. However, this prevalence of ICTs is hampered by the problem of low income levels in developing countries to the point where people cannot timeously replace or repair their ICT devices when damaged or lost; and this problem serves as a motivation for this study whose aim is to examine the perceptions of teachers on the reliability of cellphones when used for teaching and learning purposes. The research objectives unfolding this aim are of two types: objectives on the selection and design of theories and models, and objectives on the empirical testing of these theories and models. The first type of objectives is achieved using content analysis in an extensive literature survey, and the second type of objectives is achieved through a survey of high school teachers from the ILembe and Umgungudlovu districts in the KwaZuluNatal province of South Africa. Data collected from this questionnaire based survey is analysed in SPSS using descriptive statistics and Pearson correlations after checking the reliability and validity of the questionnaire. The main hypothesis driving this study is that there is a relationship between the demographics and the attribution identity of teachers on one hand, and their perceptions on the reliability of cellphones on the other hand, as suggested by existing literature; except that attribution identities are considered in this study under three angles: intention, knowledge and ability, and action. The results of this study confirm that the perceptions of teachers on the reliability of cellphones for teaching and learning are affected by the school location of these teachers, and by their perceptions on learners’ cellphones usage intentions and actual use.

Keywords: attribution, cellphones, e-learning, reliability

Procedia PDF Downloads 401

Authors: John Okanda Okwaro

Abstract:

Carbon (IV) oxide (CO₂) is emitted majorly from fossil fuel combustion and industrial production. The sources of interest of carbon (IV) oxide in the study area are mining activities, transport systems, and industrial processes. This study is aimed at building models that will help in monitoring the emissions within the study area. Three scenarios were discussed, namely: pessimistic scenario, business-as-usual scenario, and optimistic scenario. The result showed that there was a reduction in carbon dioxide concentration by approximately 50.5 ppm between March 2020 and January 2021 inclusive. This is majorly due to reduced human activities that led to decreased consumption of energy. Also, the CO₂ concentration trend follows the business-as-usual scenario (BAU) path. From the models, the pessimistic, business-as-usual, and optimistic scenarios give CO₂ concentration of about 545.9 ppm, 408.1 ppm, and 360.1 ppm, respectively, on December 31st, 2021. This research helps paint the picture to the policymakers of the relationship between energy sources and CO₂ emissions. Since the reduction in CO₂ emission was due to decreased use of fossil fuel as there was a decrease in economic activities, then if Kenya relies more on green energy than fossil fuel in the post-COVID-19 period, there will be more CO₂ emission reduction. That is, the CO₂ concentration trend is likely to follow the optimistic scenario path, hence a reduction in CO₂ concentration of about 48 ppm by the end of the year 2021. This research recommends investment in solar energy by energy-intensive companies, mine machinery and equipment maintenance, investment in electric vehicles, and doubling tree planting efforts to achieve the 10% cover.

Keywords: forecasting, greenhouse gas, green energy, hierarchical data format

Procedia PDF Downloads 166

Authors: Nishi Bhuvandas, P. V. Timbadiya, P. L. Patel, P. D. Porey

Abstract:

Recent perceived climate variability raises concerns with unprecedented hydrological phenomena and extremes. Distribution and circulation of the waters of the Earth become increasingly difficult to determine because of additional uncertainty related to anthropogenic emissions. According to the sixth Intergovernmental Panel on Climate Change (IPCC) Technical Paper on Climate Change and water, changes in the large-scale hydrological cycle have been related to an increase in the observed temperature over several decades. Although many previous research carried on effect of change in climate on hydrology provides a general picture of possible hydrological global change, new tools and frameworks for modelling hydrological series with nonstationary characteristics at finer scales, are required for assessing climate change impacts. Of the downscaling techniques, dynamic downscaling is usually based on the use of Regional Climate Models (RCMs), which generate finer resolution output based on atmospheric physics over a region using General Circulation Model (GCM) fields as boundary conditions. However, RCMs are not expected to capture the observed spatial precipitation extremes at a fine cell scale or at a basin scale. Statistical downscaling derives a statistical or empirical relationship between the variables simulated by the GCMs, called predictors, and station-scale hydrologic variables, called predictands. The main focus of the paper is on the need for using statistical downscaling techniques for projection of local hydrometeorological variables under climate change scenarios. The projections can be then served as a means of input source to various hydrologic models to obtain streamflow, evapotranspiration, soil moisture and other hydrological variables of interest.

Keywords: climate change, downscaling, GCM, RCM

Procedia PDF Downloads 405

Authors: Midiya Khademi, Ali Nikoo, Shabnam Rahimnezhad Baghche Jooghi

Abstract:

Cardiovascular diseases (CVDs) are the main cause of death globally. Most CVDs can be prevented by avoiding habitual risk factors. Separate from the habitual risk factors, there are some inherent factors in each individual that can increase the risk potential of CVDs. Vessel shapes and geometry are influential factors, having great impact on the blood flow and the hemodynamic behavior of the vessels. In the present study, the influence of bifurcation angle on blood flow characteristics is studied. In order to approach this topic, by simplifying the details of the bifurcation, three models with angles 30°, 45°, and 60° were created, then by using CFD analysis, the response of these models for stable flow and pulsatile flow was studied. In the conducted simulation in order to eliminate the influence of other geometrical factors, only the angle of the bifurcation was changed and other parameters remained constant during the research. Simulations are conducted under dynamic and stable condition. In the stable flow simulation, a steady velocity of 0.17 m/s at the inlet plug was maintained and in dynamic simulations, a typical LAD flow waveform is implemented. The results show that the bifurcation angle has an influence on the maximum speed of the flow. In the stable flow condition, increasing the angle lead to decrease the maximum flow velocity. In the dynamic flow simulations, increasing the bifurcation angle lead to an increase in the maximum velocity. Since blood flow has pulsatile characteristics, using a uniform velocity during the simulations can lead to a discrepancy between the actual results and the calculated results.

Keywords: coronary artery, cardiovascular disease, bifurcation, atherosclerosis, CFD, artery wall shear stress

Procedia PDF Downloads 163

Authors: B. R. Campomanes-Alvarez, P. Quiros, B. Fernandez

Abstract:

Automatic Speech Recognition (ASR) is a machine-based process of decoding and transcribing oral speech. A typical ASR system receives acoustic input from a speaker or an audio file, analyzes it using algorithms, and produces an output in the form of a text. Some speech recognition systems use Hidden Markov Models (HMMs) to deal with the temporal variability of speech and Gaussian Mixture Models (GMMs) to determine how well each state of each HMM fits a short window of frames of coefficients that represents the acoustic input. Another way to evaluate the fit is to use a feed-forward neural network that takes several frames of coefficients as input and produces posterior probabilities over HMM states as output. Deep neural networks (DNNs) that have many hidden layers and are trained using new methods have been shown to outperform GMMs on a variety of speech recognition systems. Acoustic models for state-of-the-art ASR systems are usually training on massive amounts of data. However, audio files with their corresponding transcriptions can be difficult to obtain, especially in the Spanish language. Hence, in the case of these low-resource scenarios, building an ASR model is considered as a complex task due to the lack of labeled data, resulting in an under-trained system. Semi-supervised learning approaches arise as necessary tasks given the high cost of transcribing audio data. The main goal of this proposal is to develop a procedure based on acoustic semi-supervised learning for Spanish ASR systems by using DNNs. This semi-supervised learning approach consists of: (a) Training a seed ASR model with a DNN using a set of audios and their respective transcriptions. A DNN with a one-hidden-layer network was initialized; increasing the number of hidden layers in training, to a five. A refinement, which consisted of the weight matrix plus bias term and a Stochastic Gradient Descent (SGD) training were also performed. The objective function was the cross-entropy criterion. (b) Decoding/testing a set of unlabeled data with the obtained seed model. (c) Selecting a suitable subset of the validated data to retrain the seed model, thereby improving its performance on the target test set. To choose the most precise transcriptions, three confidence scores or metrics, regarding the lattice concept (based on the graph cost, the acoustic cost and a combination of both), was performed as selection technique. The performance of the ASR system will be calculated by means of the Word Error Rate (WER). The test dataset was renewed in order to extract the new transcriptions added to the training dataset. Some experiments were carried out in order to select the best ASR results. A comparison between a GMM-based model without retraining and the DNN proposed system was also made under the same conditions. Results showed that the semi-supervised ASR-model based on DNNs outperformed the GMM-model, in terms of WER, in all tested cases. The best result obtained an improvement of 6% relative WER. Hence, these promising results suggest that the proposed technique could be suitable for building ASR models in low-resource environments.

Keywords: automatic speech recognition, deep neural networks, machine learning, semi-supervised learning

Procedia PDF Downloads 339

Authors: Hojjat Mohammadi, Randall Divito, Gary J. E. Kramer

Abstract:

Tunnelling and pipe jacking in very soft soils (fat clays), even with an Earth Pressure Balance tunnel boring machine (EPBM), can cause large ground displacements. In this study, the short-term and long-term ground and tunnel response is predicted for twin, pipe-jacked EPBM 3 meter diameter tunnels with a narrow pillar width. Initial modelling indicated complete closure of the annulus gap at the tail shield onto the centrifugally cast, glass-fiber-reinforced, polymer mortar jacking pipe (FRP). Numerical modelling was employed to simulate the excavation and support installation sequence, examine the ground response during excavation, confirm the adequacy of the pillar width and check the structural adequacy of the installed pipe. In the numerical models, Mohr-Coulomb constitutive model with the effect of unloading was adopted for the fat clays, while for the bedrock layer, the generalized Hoek-Brown was employed. The numerical models considered explicit excavation sequences and different levels of ground convergence prior to support installation. The well-studied excavation sequences made the analysis possible for this study on a very soft clay, otherwise, obtaining the convergency in the numerical analysis would be impossible. The predicted results indicate that the ground displacements around the tunnel and its effect on the pipe would be acceptable despite predictions of large zones of plastic behaviour around the tunnels and within the entire pillar between them due to excavation-induced ground movements.

Keywords: finite element modeling (FEM), pipe-jacked tunneling, very soft clay, EPBM

Procedia PDF Downloads 80

Authors: Solmaz Jahed Shiran, John Hearne, Tayebeh Saghapour

Abstract:

Daily travel behavior is strongly influenced by the location of the places of residence, education, and employment. Hence a change in those locations due to a move or changes in an occupation leads to a change in travel behavior. Given the interventions of housing mobility and travel behaviors, the hypothesis is that a mobile housing market allows households to move as a result of any change in their life course, allowing them to be closer to central services, public transport facilities and workplace and hence reducing the time spent by individuals on daily travel. Conversely, household’s immobility may lead to longer commutes of residents, for example, after a change of a job or a need for new services such as schools for children who have reached their school age. This paper aims to investigate the association between residential mobility and travel behavior. The Victorian Integrated Survey of Travel and Activity (VISTA) data is used for the empirical analysis. Car ownership and journey to work time and distance of employed people are used as indicators of travel behavior. Change of usual residence within the last five years used to identify movers and non-movers. Statistical analysis, including regression models, is used to compare the travel behavior of movers and non-movers. The results show travel time, and the distance does not differ for movers and non-movers. However, this is not the case when taking into account the residence tenure-type. In addition, car ownership rate and number found to be significantly higher for non-movers. It is hoped that the results from this study will contribute to a better understanding of factors other than common socioeconomic and built environment features influencing travel behavior.

Keywords: journey to work, regression models, residential mobility, commute mode, car ownership

Procedia PDF Downloads 133

Authors: Matthias Stange

Abstract:

Architectural visualization has been practiced for as long as there have been buildings. Visualization (lat.: visibilis "visible") generally refers to bringing abstract data and relationships into a graphically, visually comprehensible form. Particularly, visualization refers to the process of translating relationships that are difficult to formulate linguistically or logically into visual media (e.g., drawings or models) to make them comprehensible. Building owners have always been interested in knowing how their building will look before it is built. In the empirical part of this study, the roots of architectural visualization are examined, starting from the ancient civilizations to the end of the Roman Empire. Extensive literature research on architectural theory and architectural history forms the basis for this analysis. The focus of the analysis is basic research from the emergence of the first two-dimensional drawings in the Neolithic period to the triggers of significant further developments of architectural representation, as well as their importance for subsequent methods and the transmission of knowledge over the following epochs. The analysis focuses on the development of analog methods of representation from the first Neolithic house floor plans to the Greek detailed stone models and paper drawings in the Roman Empire. In particular, the question of socio-cultural, socio-political, and economic changes as possible triggers for the development of representational media and methods will be analyzed. The study has shown that the development of visual building representation has been driven by scientific, technological, and social developments since the emergence of the first civilizations more than 6000 years ago first by the change in human’s subsistence strategy, from food appropriation by hunting and gathering to food production by agriculture and livestock, and the sedentary lifestyle required for this.

Keywords: ancient Greece, ancient orient, Roman Empire, architectural visualization

Procedia PDF Downloads 115

Authors: Dicko Ali Hamadi, Tong-Yette Nicolas, Gilles Benjamin, Faure Francois, Palombi Olivier

Abstract:

A novel hybrid model of the lumbar spine, allowing fast static and dynamic simulations of the disc pressure and the spine mobility, is introduced in this work. Our contribution is to combine rigid bodies, deformable finite elements, articular constraints, and springs into a unique model of the spine. Each vertebra is represented by a rigid body controlling a surface mesh to model contacts on the facet joints and the spinous process. The discs are modeled using a heterogeneous tetrahedral finite element model. The facet joints are represented as elastic joints with six degrees of freedom, while the ligaments are modeled using non-linear one-dimensional elastic elements. The challenge we tackle is to make these different models efficiently interact while respecting the principles of Anatomy and Mechanics. The mobility, the intradiscal pressure, the facet joint force and the instantaneous center of rotation of the lumbar spine are validated against the experimental and theoretical results of the literature on flexion, extension, lateral bending as well as axial rotation. Our hybrid model greatly simplifies the modeling task and dramatically accelerates the simulation of pressure within the discs, as well as the evaluation of the range of motion and the instantaneous centers of rotation, without penalizing precision. These results suggest that for some types of biomechanical simulations, simplified models allow far easier modeling and faster simulations compared to usual full-FEM approaches without any loss of accuracy.

Keywords: hybrid, modeling, fast simulation, lumbar spine

Procedia PDF Downloads 305

Authors: Juan Sebastián Rivera Soriano, Rosa Urbano Gutiérrez

Abstract:

The Shodan House and the Sarabhai House (Ahmedabad, India, 1954 and 1955, respectively) are considered some of the most important works of Le Corbusier produced in the last stage of his career. There are some academic publications that study the compositional and formal aspects of their architectural design, but there is no in-depth investigation into how the climatic conditions of this region were a determining factor in the design decisions implemented in these projects. This paper argues that Le Corbusier developed a specific architectural design strategy for these buildings based on scientific research on climate in the Indian context. This new language was informed by a pioneering study and interpretation of climatic data as a design methodology that would even involve the development of new design tools. This study investigated whether their use of climatic data meets values and levels of accuracy obtained with contemporary instruments and tools, such as Energy Plus weather data files and Climate Consultant. It also intended to find out if Le Corbusier's office’s intentions and decisions were indeed appropriate and efficient for those climate conditions by assessing these projects using BIM models and energy performance simulations from Design Builder. Accurate models were built using original historical data through archival research. The outcome is to provide a new understanding of the environment of these houses through the combination of modern building science and architectural history. The results confirm that in these houses, it was achieved a model of low energy consumption. This paper contributes new evidence not only on exemplary modern architecture concerned with environmental performance but also on how it developed progressive thinking in this direction.

Keywords: bioclimatic architecture, Le Corbusier, Shodan, Sarabhai Houses

Procedia PDF Downloads 63

Authors: Mustafa Sertçelik, Turan Çalban, Hacali Necefoğlu, Sabri Çolak

Abstract:

In this study, Chevreul’s salt solubility and its dissolution kinetics in ammonium chloride solutions were investigated. Chevreul’s salt that we used in the studies was obtained by using the optimum conditions (ammonium sulphide concentration; 0,4 M, copper sulphate concentration; 0,25 M, temperature; 60°C, stirring speed; 600 rev/min, pH; 4 and reaction time; 15 mins) determined by T. Çalban et al. Chevreul’s salt solubility in ammonium chloride solutions and the kinetics of dissolution were investigated. The selected parameters that affect solubility were reaction temperature, concentration of ammonium chloride, stirring speed, and solid/liquid ratio. Correlation of experimental results had been achieved using linear regression implemented in the statistical package program statistica. The effect of parameters on Chevreul’s salt solubility was examined and integrated rate expression of dissolution rate was found using kinetic models in solid-liquid heterogeneous reactions. The results revealed that the dissolution rate of Chevreul’s salt was decreasing while temperature, concentration of ammonium chloride and stirring speed were increasing. On the other hand, dissolution rate was found to be decreasing with the increase of solid/liquid ratio. Based on result of the applications of the obtained experimental results to the kinetic models, we can deduce that Chevreul’s salt dissolution rate is controlled by diffusion through the ash (or product layer). Activation energy of the reaction of dissolution was found as 74.83 kJ/mol. The integrated rate expression along with the effects of parameters on Chevreul's salt solubility was found to be as follows: 1-3(1-X)2/3+2(1-X)= [2,96.1013.(CA)3,08 .(S/L)-038.(W)1,23 e-9001,2/T].t

Keywords: Chevreul's salt, copper, ammonium chloride, ammonium sulphide, dissolution kinetics

Procedia PDF Downloads 306

Authors: E. Cascarosa-Salillas, J. Pozuelo-Muñoz, C. Rodríguez-Casals, A. de Echave

Abstract:

This study focuses on improving the understanding of the capillarity phenomenon among Primary and Secondary Education students. Despite being a common concept in daily life and covered in various subjects, students’ comprehension remains limited. This work explores inquiry-based teaching methods to build a conceptual foundation of capillarity by examining the forces involved. The study adopts an inquiry-based teaching approach supported by research emphasizing the importance of modeling in science education. Scientific modeling aids students in applying knowledge across varied contexts and developing systemic thinking, allowing them to construct scientific models applicable to everyday situations. This methodology fosters the development of scientific competencies such as observation, hypothesis formulation, and communication. The research was structured as a case study with activities designed for Spanish Primary and Secondary Education students aged 9 to 13. The process included curriculum analysis, the design of an activity sequence, and its implementation in classrooms. Implementation began with questions that students needed to resolve using available materials, encouraging observation, experimentation, and the re-contextualization of activities to everyday phenomena where capillarity is observed. Data collection tools included audio and video recordings of the sessions, which were transcribed and analyzed alongside the students' written work. Students' drawings on capillarity were also collected and categorized. Qualitative analyses of the activities showed that, through inquiry, students managed to construct various models of capillarity, reflecting an improved understanding of the phenomenon. Initial activities allowed students to express prior ideas and formulate hypotheses, which were then refined and expanded in subsequent sessions. The generalization and use of graphical representations of their ideas on capillarity, analyzed alongside their written work, enabled the categorization of capillarity models: Intuitive Model: A visual and straightforward representation without explanations of how or why it occurs. Simple symbolic elements, such as arrows to indicate water rising, are used without detailed or causal understanding. It reflects an initial, immediate perception of the phenomenon, interpreted as something that happens "on its own" without delving into the microscopic level. Explanatory Intuitive Model: Students begin to incorporate causal explanations, though still limited and without complete scientific accuracy. They represent the role of materials and use basic terms such as ‘absorption’ or ‘attraction’ to describe the rise of water. This model shows a more complex understanding where the phenomenon is not only observed but also partially explained in terms of interaction, though without microscopic detail. School Scientific Model: This model reflects a more advanced and detailed understanding. Students represent the phenomenon using specific scientific concepts like ‘surface tension,’ cohesion,’ and ‘adhesion,’ including structured explanations connecting microscopic and macroscopic levels. At this level, students model the phenomenon as a coherent system, demonstrating how various forces or properties interact in the capillarity process, with representations on a microscopic level. The study demonstrated that the capillarity phenomenon can be effectively approached in class through the experimental observation of everyday phenomena, explained through guided inquiry learning. The methodology facilitated students’ construction of capillarity models and served to analyze an interaction phenomenon of different forces occurring at the microscopic level.

Keywords: capillarity, inquiry-based learning, scientific modeling, primary and secondary education, conceptual understanding, Drawing analysis.

Procedia PDF Downloads 12