Search results for: prediction model accuracy

Authors: Luana Conte, Giorgio De Nunzio, Giuseppe Raso, Donato Cascio

Abstract:

The intestinal epithelium serves as a pivotal model for studying stem cell biology and diseases such as colorectal cancer. Intestinal epithelial organoids, which replicate many in vivo features of the intestinal epithelium, are increasingly used as research models. However, manual classification of organoids is labor-intensive and prone to subjectivity, limiting scalability. In this study, we developed an automated object-detection algorithm to classify intestinal organoids in transmitted-light microscopy images. Our approach utilizes the YOLOv10 medium model (YOLO10m), a state-of-the-art object-detection algorithm, to predict and classify objects within labeled bounding boxes. The model was fine-tuned on a publicly available dataset containing 840 manually annotated images with 23,066 total annotations, averaging 28.2 annotations per image (median: 21; range: 1–137). It was trained to identify four categories: cysts, early organoids, late organoids, and spheroids, using a 90:10 train-validation split over 150 epochs. Model performance was assessed using mean average precision (mAP), precision, and recall metrics. The mAP, a standard metric ranging from 0 to 1 (with 1 indicating perfect agreement with manual labeling), was calculated at a 50% overlap threshold (mAP=0.5). Optimal performance was achieved at epoch 80, with an mAP of 0.85, precision of 0.78, and recall of 0.80 on the validation dataset. Classspecific mAP values were highest for cysts (0.87), followed by late organoids (0.83), early organoids (0.76), and spheroids (0.68). Additionally, the model demonstrated the ability to measure organoid sizes and classify them with accuracy comparable to expert scientists, while operating significantly faster. This automated pipeline represents a robust tool for large-scale, high-throughput analysis of intestinal organoids, paving the way for more efficient research in organoid biology and related fields.

Keywords: intestinal organoids, object detection, YOLOv10, transmitted-light microscopy

Procedia PDF Downloads 12

Authors: Usama Ahmed

Abstract:

Data mining is the process of analyze data which are used to predict helpful information. It is the field of research which solve various type of problem. In data mining, classification is an important technique to classify different kind of data. Diabetes is most common disease. This paper implements different classification technique using Waikato Environment for Knowledge Analysis (WEKA) on diabetes dataset and find which algorithm is suitable for working. The best classification algorithm based on diabetic data is Naïve Bayes. The accuracy of Naïve Bayes is 76.31% and take 0.06 seconds to build the model.

Keywords: data mining, classification, diabetes, WEKA

Procedia PDF Downloads 152

Authors: Aziz Ahmad, Guillermo Ramirez-Prado

Abstract:

Governments around the world encourage the use of solar water heating in residential houses due to the low maintenance requirements and efficiency of the solar collector water heating systems. The aim of this work is to study a domestic solar water heating system in a residential building to develop a model of the entire solar water heating system including flat-plate solar collector and storage tank. The proposed model is adaptable to any households and location. The model can be used to test different types of controllers and can provide efficiency as well as economic analysis. The proposed model is based on the heat and mass transfer equations along with assumptions applied in the model which can be modified for a variety of different solar water heating systems and sizes. Simulation results of the model were compared with the actual system which shows similar trends.

Keywords: solar thermal systems, solar water heating, solar collector model, hot water tank model, solar controllers

Procedia PDF Downloads 277

Authors: Chakib Jerry, Mounir Jerry

Abstract:

In this paper, a mathematical model of infested honey bees colonies is formulated in order to investigate Colony Collapse Disorder in a honeybee colony. CCD, as it is known, is a major problem on honeybee farms because of the massive decline in colony numbers. We introduce to the model a control variable which represents forager protection. We study the controlled model to derive conditions under which the bee colony can fight off epidemic. Secondly we study the problem of minimizing prevention cost under model’s dynamics constraints.

Keywords: honey bee, disease transmission model, disease control honeybees, optimal control

Procedia PDF Downloads 431

Authors: Yeon-Jong Jeong, Jun-Hong Park, Hyuk Choi

Abstract:

In this study, we have studied the design of intermediate die in multipass drawing. Research has been continuously studied because of the advantage of better dimensional accuracy, smooth surface and improved mechanical properties in the case of drawing. Among them, multipass drawing, which is a method to realize complicated shape by drawing, was discussed in this study. The most important factor in the multipass drawing is the dimensional accuracy and simplify the process. To accomplish this, a multistage shape drawing was performed using various intermediate die shape designs, and finite element analysis was performed.

Keywords: FEM (Finite Element Method), multipass drawing, intermediate die, hollow pipe

Procedia PDF Downloads 317

Authors: Mahdieh Khalilinezhad, Silvana Dellepiane, Gianni Vernazza

Abstract:

The aim of the present work is to build a model based on tissue characterization that is able to discriminate pathological and non-pathological regions from three-phasic CT images. Based on feature selection in different phases, in this research, we design a neural network system that has optimal neuron number in a hidden layer. Our approach consists of three steps: feature selection, feature reduction, and classification. For each ROI, 6 distinct set of texture features are extracted such as first order histogram parameters, absolute gradient, run-length matrix, co-occurrence matrix, autoregressive model, and wavelet, for a total of 270 texture features. We show that with the injection of liquid and the analysis of more phases the high relevant features in each region changed. Our results show that for detecting HCC tumor phase3 is the best one in most of the features that we apply to the classification algorithm. The percentage of detection between these two classes according to our method, relates to first order histogram parameters with the accuracy of 85% in phase 1, 95% phase 2, and 95% in phase 3.

Keywords: multi-phasic liver images, texture analysis, neural network, hidden layer

Procedia PDF Downloads 264

Authors: T. Benazzouz, K. Auhmani

Abstract:

This paper presents a Hospital 4.0 Maturity Assessment Model based on the Industry 4.0 concepts. The self-assessment model defines current and target states of digital transformation by considering multiple aspects of a hospital and a healthcare supply chain. The developed model was validated and evaluated on real-life cases. The resulting model consisted of 5 domains: Technology, Strategy 4.0, Human resources 4.0 & Culture 4.0, Supply chain 4.0 management, and Patient journeys management. Each domain is further divided into several sub-domains, totally 34 sub-domains are identified, that reflect different facets of a hospital 4.0 mature organization.

Keywords: hospital 4.0, Industry 4.0, maturity assessment model, supply chain 4.0, patient

Procedia PDF Downloads 99

Authors: H. Abas Nejad, A. R. Teymoori

Abstract:

Facial expression recognition is one of the open problems in computer vision. Robust neutral face recognition in real time is a major challenge for various supervised learning based facial expression recognition methods. This is due to the fact that supervised methods cannot accommodate all appearance variability across the faces with respect to race, pose, lighting, facial biases, etc. in the limited amount of training data. Moreover, processing each and every frame to classify emotions is not required, as the user stays neutral for the majority of the time in usual applications like video chat or photo album/web browsing. Detecting neutral state at an early stage, thereby bypassing those frames from emotion classification would save the computational power. In this work, we propose a light-weight neutral vs. emotion classification engine, which acts as a preprocessor to the traditional supervised emotion classification approaches. It dynamically learns neutral appearance at Key Emotion (KE) points using a textural statistical model, constructed by a set of reference neutral frames for each user. The proposed method is made robust to various types of user head motions by accounting for affine distortions based on a textural statistical model. Robustness to dynamic shift of KE points is achieved by evaluating the similarities on a subset of neighborhood patches around each KE point using the prior information regarding the directionality of specific facial action units acting on the respective KE point. The proposed method, as a result, improves ER accuracy and simultaneously reduces the computational complexity of ER system, as validated on multiple databases.

Keywords: neutral vs. emotion classification, Constrained Local Model, procrustes analysis, Local Binary Pattern Histogram, statistical model

Procedia PDF Downloads 343

Authors: Eiman Kattan, Hong Wei

Abstract:

In using Convolutional Neural Network (CNN) for classification, there is a set of hyperparameters available for the configuration purpose. This study aims to evaluate the impact of a range of parameters in CNN architecture i.e. AlexNet on land cover classification based on four remotely sensed datasets. The evaluation tests the influence of a set of hyperparameters on the classification performance. The parameters concerned are epoch values, batch size, and convolutional filter size against input image size. Thus, a set of experiments were conducted to specify the effectiveness of the selected parameters using two implementing approaches, named pertained and fine-tuned. We first explore the number of epochs under several selected batch size values (32, 64, 128 and 200). The impact of kernel size of convolutional filters (1, 3, 5, 7, 10, 15, 20, 25 and 30) was evaluated against the image size under testing (64, 96, 128, 180 and 224), which gave us insight of the relationship between the size of convolutional filters and image size. To generalise the validation, four remote sensing datasets, AID, RSD, UCMerced and RSCCN, which have different land covers and are publicly available, were used in the experiments. These datasets have a wide diversity of input data, such as number of classes, amount of labelled data, and texture patterns. A specifically designed interactive deep learning GPU training platform for image classification (Nvidia Digit) was employed in the experiments. It has shown efficiency in both training and testing. The results have shown that increasing the number of epochs leads to a higher accuracy rate, as expected. However, the convergence state is highly related to datasets. For the batch size evaluation, it has shown that a larger batch size slightly decreases the classification accuracy compared to a small batch size. For example, selecting the value 32 as the batch size on the RSCCN dataset achieves the accuracy rate of 90.34 % at the 11th epoch while decreasing the epoch value to one makes the accuracy rate drop to 74%. On the other extreme, setting an increased value of batch size to 200 decreases the accuracy rate at the 11th epoch is 86.5%, and 63% when using one epoch only. On the other hand, selecting the kernel size is loosely related to data set. From a practical point of view, the filter size 20 produces 70.4286%. The last performed image size experiment shows a dependency in the accuracy improvement. However, an expensive performance gain had been noticed. The represented conclusion opens the opportunities toward a better classification performance in various applications such as planetary remote sensing.

Keywords: CNNs, hyperparamters, remote sensing, land cover, land use

Procedia PDF Downloads 173

Authors: P. Nafisi Poor, P. Javid

Abstract:

Over the past few years, the idea of adaptive buildings and specifically, adaptive building energy management systems (ABEMS) has become popular. Well-performed management in terms of energy is to create a balance between energy consumption and user comfort; therefore, in new energy management models, efficient energy consumption is not the sole factor and the user's comfortability is also considered in the calculations. One of the main ways of measuring this factor is by analyzing user feedback on the conditions to understand whether they are satisfied with conditions or not. This paper provides a comprehensive review of recent approaches towards energy management systems based on users' feedbacks and subsequently performs a comparison between them premised upon their efficiency and accuracy to understand which approaches were more accurate and which ones resulted in a more efficient way of minimizing energy consumption while maintaining users' comfortability. It was concluded that the highest accuracy rate among the presented works was 95% accuracy in determining satisfaction and up to 51.08% energy savings can be achieved without disturbing user’s comfort. Considering the growing interest in designing and developing adaptive buildings, these studies can support diverse inquiries about this subject and can be used as a resource to support studies and researches towards efficient energy consumption while maintaining the comfortability of users.

Keywords: adaptive buildings, energy efficiency, intelligent buildings, user comfortability

Procedia PDF Downloads 137

Authors: Carolina Senabre, Sergio Valero, Miguel Lopez, Antonio Gabaldon

Abstract:

This paper presents a comprehensive survey of the short-term load forecasting (STLF). Since the behavior of consumers and producers continue changing as new technologies, it is an ongoing process, and moreover, new policies become available. The results of a research study for the Spanish Transport System Operator (REE) is presented in this paper. It is presented the improvement of the forecasting accuracy in the Balearic Islands considering the introduction of meteorological variables, such as temperature to reduce forecasting error. Variables analyzed for the forecasting in terms of overall accuracy are cloudiness, solar radiation, and wind velocity. It has also been analyzed the type of days to be considered in the research.

Keywords: short-term load forecasting, power demand, neural networks, load forecasting

Procedia PDF Downloads 194

Authors: Xudong Guan, Ainong Li, Gaohuan Liu, Chong Huang, Wei Zhao

Abstract:

Due to the low spatial resolution of MODIS data, the accuracy of small-area plaque extraction with a high degree of landscape fragmentation is greatly limited. To this end, the study combines Landsat data with higher spatial resolution and MODIS data with higher temporal resolution for decision-level fusion. Considering the importance of the land heterogeneity factor in the fusion process, it is superimposed with the weighting factor, which is to linearly weight the Landsat classification result and the MOIDS classification result. Three levels were used to complete the process of data fusion, that is the pixel of MODIS data, the pixel of Landsat data, and objects level that connect between these two levels. The multilevel decision fusion scheme was tested in two sites of the lower Mekong basin. We put forth a comparison test, and it was proved that the classification accuracy was improved compared with the single data source classification results in terms of the overall accuracy. The method was also compared with the two-level combination results and a weighted sum decision rule-based approach. The decision fusion scheme is extensible to other multi-resolution data decision fusion applications.

Keywords: image classification, decision fusion, multi-temporal, remote sensing

Procedia PDF Downloads 127

Authors: Engy S. El-Kayal, Mohamed M. S. Arafa

Abstract:

There are various causes of ankle pain including traumatic and non-traumatic causes. Various imaging techniques are available for assessment of AP. MRI is considered to be the imaging modality of choice for ankle joint evaluation with an advantage of its high spatial resolution, multiplanar capability, hence its ability to visualize small complex anatomical structures around the ankle. However, the high costs and the relatively limited availability of MRI systems, as well as the relatively long duration of the examination all are considered disadvantages of MRI examination. Therefore there is a need for a more rapid and less expensive examination modality with good diagnostic accuracy to fulfill this gap. HRU has become increasingly important in the assessment of ankle disorders, with advantages of being fast, reliable, of low cost and readily available. US can visualize detailed anatomical structures and assess tendinous and ligamentous integrity. The aim of this study was to compare the diagnostic accuracy of HRU with MRI in the assessment of patients with AP. We included forty patients complaining of AP. All patients were subjected to real-time HRU and MRI of the affected ankle. Results of both techniques were compared to surgical and arthroscopic findings. All patients were examined according to a defined protocol that includes imaging the tendon tears or tendinitis, muscle tears, masses, or fluid collection, ligament sprain or tears, inflammation or fluid effusion within the joint or bursa, bone and cartilage lesions, erosions and osteophytes. Analysis of the results showed that the mean age of patients was 38 years. The study comprised of 24 women (60%) and 16 men (40%). The accuracy of HRU in detecting causes of AP was 85%, while the accuracy of MRI in the detection of causes of AP was 87.5%. In conclusions: HRU and MRI are two complementary tools of investigation with the former will be used as a primary tool of investigation and the latter will be used to confirm the diagnosis and the extent of the lesion especially when surgical interference is planned.

Keywords: ankle pain (AP), high-resolution ultrasound (HRU), magnetic resonance imaging (MRI) ultrasonography (US)

Procedia PDF Downloads 194

Authors: Huilan Yao, Huaixin Zhang

Abstract:

Simulations of the transition flow of model propellers are important for predicting hydrodynamic performance and studying scale effects. In this paper, the transition flow of a model propeller under different loadings are simulated using a transition model provided by STAR-CCM+, and the influence of turbulence intensity (TI) on the transition, especially friction and pressure components of propeller performance, was studied. Before that, the transition model was applied to simulate the transition flow of a flat plate and an airfoil. Predicted transitions agree well with experimental results. Then, the transition model was applied for propeller simulations in open water, and the influence of TI was studied. Under the heavy and moderate loadings, thrust and torque of the propeller predicted by the transition model (different TI) and two turbulence models are very close and agree well with measurements. However, under the light loading, only the transition model with low TI predicts the most accurate results. Above all, the friction components of propeller performance predicted by the transition model with different TI have obvious difference.

Keywords: transition flow, model propellers, hydrodynamic performance, numerical simulation

Procedia PDF Downloads 270

Authors: Jagadish Rao Padubidri, Mehak Bhandary, Sowmya J. Rao

Abstract:

Introduction: The significance of the human foot and its measurements in identifying an individual has been proved a lot of times by different studies in different geographical areas and its association to the stature and gender of the individual has been justified by many researches. In our study we have used different foot measurements including the length, width, malleol height and navicular height for establishing its association to stature and gender and to find out its accuracy. The purpose of this study is to show the relation of foot measurements with stature and gender, and to derive Multiple and Logistic regression equations for stature and gender estimation in South Indian population. Materials and Methods: The subjects for this study were 200 South Indian students out of which 100 were females and 100 were males, aged between 18 to 24 years. The data for the present study included the stature, foot length, foot breath, foot malleol height, foot navicular height of both right and left foot. Descriptive statistics, T-test and Pearson correlation coefficients were derived between stature, gender and foot measurements. The stature was estimated from right and left foot measurements for both male and female South Indian population using multiple regression analysis and logistic regression analysis for gender estimation. Results: The means, standard deviation, stature, right and left foot measurements and T-test in male population were higher than in females. LFL (Left foot length) is more than RFL (Right Foot length) in male groups, but in female groups the length of both foot are almost equal [RFL=226.6, LFL=227.1]. There is not much of difference in means of RFW (Right foot width) and LFW (Left foot width) in both the genders. Significant difference were seen in mean values of malleol and navicular height of right and left feet in male gender. No such difference was seen in female subjects. Conclusions: The study has successfully demonstrated the correlation of foot length in stature estimation in all the three study groups in both right and left foot. Next in parameters are Foot width and malleol height in estimating stature among male and female groups. Navicular height of both right and left foot showed poor relationship with stature estimation in both male and female groups. Multiple regression equations for both right and left foot measurements to estimate stature were derived with standard error ranging from 11-12 cm in males and 10-11 cm in females. The SEE was 5.8 when both male and female groups were pooled together. The logistic regression model which was derived to determine gender showed 85% accuracy and 92.5% accuracy using right and left foot measurements respectively. We believe that stature and gender can be estimated with foot measurements in South Indian population.

Keywords: foot length, gender, stature, South Indian

Procedia PDF Downloads 338

Authors: Mohamed Jama Madar, Oso Wilis

Abstract:

E-learning is the application of information technology in the teaching and learning process. This paper presents the Funnel model as a solution for the problems of implementation of e-learning in tertiary education institutions. While existing models such as TAM, theory-based e-learning and pedagogical model have been used over time, they have generally been found to be inadequate because of their tendencies to treat materials development, instructional design, technology, delivery and governance as separate and isolated entities. Yet it is matching components that bring framework of e-learning strategic implementation. The Funnel model enhances all these into one and applies synchronously and asynchronously to e-learning implementation where the only difference is modalities. Such a model for e-learning implementation has been lacking. The proposed Funnel model avoids ad-ad-hoc approach which has made other systems unused or inefficient, and compromised educational quality. Therefore, the proposed Funnel model should help tertiary education institutions adopt and develop effective and efficient e-learning system which meets users’ requirements.

Keywords: e-learning, pedagogical, technology, strategy

Procedia PDF Downloads 457

Authors: Muhammad Usman Akram

Abstract:

This research work based on GNSS-Aided Photogrammetry for Digital Mapping. It focuses on topographic survey of an area or site which is to be used in future Planning & development (P&D) or can be used for further, examination, exploration, research and inspection. Survey and Mapping in hard-to-access and hazardous areas are very difficult by using traditional techniques and methodologies; as well it is time consuming, labor intensive and has less precision with limited data. In comparison with the advance techniques it is saving with less manpower and provides more precise output with a wide variety of multiple data sets. In this experimentation, Aerial Photogrammetry technique is used where an UAV flies over an area and captures geocoded images and makes a Three-Dimensional Model (3-D Model), UAV operates on a user specified path or area with various parameters; Flight altitude, Ground sampling distance (GSD), Image overlapping, Camera angle etc. For ground controlling, a network of points on the ground would be observed as a Ground Control point (GCP) using Differential Global Positioning System (DGPS) in PPK or RTK mode. Furthermore, that raw data collected by UAV and DGPS will be processed in various Digital image processing programs and Computer Aided Design software. From which as an output we obtain Points Dense Cloud, Digital Elevation Model (DEM) and Ortho-photo. The imagery is converted into geospatial data by digitizing over Ortho-photo, DEM is further converted into Digital Terrain Model (DTM) for contour generation or digital surface. As a result, we get Digital Map of area to be surveyed. In conclusion, we compared processed data with exact measurements taken on site. The error will be accepted if the amount of error is not breached from survey accuracy limits set by concerned institutions.

Keywords: photogrammetry, post processing kinematics, real time kinematics, manual data inquiry

Procedia PDF Downloads 37

Authors: Cristian Crespo

Abstract:

Conditional cash transfers (CCTs) have been targeted towards the poor. Thus, their targeting assessments check whether these schemes have been allocated to low-income households or individuals. However, CCTs have more than one goal and target group. An additional goal of CCTs is to increase school enrolment. Hence, students at risk of dropping out of school also are a target group. This paper analyses whether one of the most common targeting mechanisms of CCTs, a proxy means test (PMT), is suitable to identify the poor and future school dropouts. The PMT is compared with alternative approaches that use the outputs of a predictive model of school dropout. This model was built using machine learning algorithms and rich administrative datasets from Chile. The paper shows that using machine learning outputs in conjunction with the PMT increases targeting effectiveness by identifying more students who are either poor or future dropouts. This joint targeting approach increases effectiveness in different scenarios except when the social valuation of the two target groups largely differs. In these cases, the most likely optimal approach is to solely adopt the targeting mechanism designed to find the highly valued group.

Keywords: conditional cash transfers, machine learning, poverty, proxy means tests, school dropout prediction, targeting

Procedia PDF Downloads 206

Authors: K. Shiba, T. Kaburagi, Y. Kurihara

Abstract:

With global aging, people who require care, such as people with dementia (PwD), are increasing within many developed countries. And PwDs may wander and unconsciously set foot outdoors, it may lead serious accidents, such as, traffic accidents. Here, round-the-clock monitoring by caregivers is necessary, which can be a burden for the caregivers. Therefore, an automatic wandering detection system is required when an elderly person wanders outdoors, in which case the detection system transmits a ‘moving’ followed by an ‘absence’ state. In this paper, we focus on the transition from the ‘resting’ to the ‘absence’ state, via the ‘moving’ state as one of the wandering transitions. To capture the transition of the three states, our method based on the hidden Markov model (HMM) is built. Using our method, the restraint where the ‘resting’ state and ‘absence’ state cannot be transmitted to each other is applied. To validate our method, we conducted the experiment with 10 subjects. Our results show that the method can classify three states with 0.92 accuracy.

Keywords: wander, microwave Doppler sensor, respiratory frequency band, the state transition, hidden Markov model (HMM).

Procedia PDF Downloads 187

Authors: Rui Liu, Klaus Greve

Abstract:

The navigation information is a kind of dynamic geographic information, and the navigation information system is a kind of special geographic information system. At present, there are many researches on the application of centralized management and cross-integration application of basic geographic information. However, the idea of information integration and sharing is not deeply applied into the research of navigation information service. And the imperfection of navigation target coordination and navigation information sharing mechanism under certain navigation tasks has greatly affected the reliability and scientificity of navigation service such as path planning. Considering this, the project intends to study the multi-source information fusion and multi-objective united navigation information interaction mechanism: first of all, investigate the actual needs of navigation users in different areas, and establish the preliminary navigation information classification and importance level model; and then analyze the characteristics of the remote sensing and GIS vector data, and design the fusion algorithm from the aspect of improving the positioning accuracy and extracting the navigation environment data. At last, the project intends to analyze the feature of navigation information of the land, sea and air navigation targets, and design the united navigation data standard and navigation information sharing model under certain navigation tasks, and establish a test navigation system for united navigation simulation experiment. The aim of this study is to explore the theory of united navigation service and optimize the navigation information service model, which will lay the theory and technology foundation for the united navigation of land, sea and air targets.

Keywords: information fusion, united navigation, dynamic path planning, navigation information visualization

Procedia PDF Downloads 292

Authors: Venkata Narayana Chejarla, Nafisa Parvez Shaik, Gopi Vara Prasad Marabathula, Deva Kumar Bejjam

Abstract:

Yoga is an advanced, well-recognized method with roots in Indian philosophy. Yoga benefits both the body and the psyche. Yoga is a regular exercise that helps people relax and sleep better while also enhancing their balance, endurance, and concentration. Yoga can be learned in a variety of settings, including at home with the aid of books and the internet as well as in yoga studios with the guidance of an instructor. Self-learning does not teach the proper yoga poses, and doing them without the right instruction could result in significant injuries. We developed "Online Yoga Asana Trainer using Deep Learning" so that people could practice yoga without a teacher. Our project is developed using Tensorflow, Movenet, and Keras models. The system makes use of data from Kaggle that includes 25 different yoga poses. The first part of the process involves applying the movement model for extracting the 17 key points of the body from the dataset, and the next part involves preprocessing, which includes building a pose classification model using neural networks. The system scores a 98.3% accuracy rate. The system is developed to work with live videos.

Keywords: yoga, deep learning, movenet, tensorflow, keras, CNN

Procedia PDF Downloads 248

Authors: Gabriela Silva-Maceda, Silvia Romero-Contreras

Abstract:

Spanish is a more transparent language than English given that it has more direct correspondences between sounds and letters. It has become important to understand how decoding and linguistic comprehension contribute to reading comprehension in the framework of the widely known Simple View Model. This study aimed to identify the level of prediction by these two components in a sample of 1st to 4th grade children attending two schools in central Mexico (one public and one private). Within each school, ten children were randomly selected in each grade level, and their parents were asked about reading habits and socioeconomic information. In total, 79 children completed three standardized tests measuring decoding (pseudo-word reading), linguistic comprehension (understanding of paragraphs) and reading comprehension using subtests from the Clinical Evaluation of Language Fundamentals-Spanish, Fourth Edition, and the Test de Lectura y Escritura en Español (LEE). The data were analyzed using hierarchical regression, with decoding as a first step and linguistic comprehension as a second step. Results showed that decoding accounted for 19.2% of the variance in reading comprehension, while linguistic comprehension accounted for an additional 10%, adding up to 29.2% of variance explained: F (2, 75)= 15.45, p <.001. Socioeconomic status derived from parental questionnaires showed a statistically significant association with the type of school attended, X2 (3, N= 79) = 14.33, p =.002. Nonetheless when analyzing the Simple View components, only decoding differences were statistically significant (t = -6.92, df = 76.81, p < .001, two-tailed); reading comprehension differences were also significant (t = -3.44, df = 76, p = .001, two-tailed). When socioeconomic status was included in the model, it predicted a 5.9% unique variance, even when already accounting for Simple View components, adding to a 35.1% total variance explained. This three-predictor model was also significant: F (3, 72)= 12.99, p <.001. In addition, socioeconomic status was significantly correlated with the amount of non-textbook books parents reported to have at home for both adults (rho = .61, p<.001) and children (rho= .47, p<.001). Results converge with a large body of literature finding socioeconomic differences in reading comprehension; in addition this study suggests that these differences were also present in decoding skills. Although linguistic comprehension differences between schools were expected, it is argued that the test used to collect this variable was not sensitive to linguistic differences, since it came from a test to diagnose clinical language disabilities. Even with this caveat, results show that the components of the Simple View Model can predict less than a third of the variance in reading comprehension in Spanish. However, the results also suggest that a fuller model of reading comprehension is obtained when considering the family’s socioeconomic status, given the potential differences shown by the socioeconomic status association with books at home, factors that are particularly important in countries where inequality gaps are relatively large.

Keywords: decoding, linguistic comprehension, reading comprehension, simple view model, socioeconomic status, Spanish

Procedia PDF Downloads 333

Authors: Basiliyos Tilahun Betru, Charles Awono Onana, Bernabe Batchakui

Abstract:

Recommender systems have become increasingly popular in recent years, and are utilized in numerous areas. Nowadays many web services provide several information for users and recommender systems have been developed as critical element of these web applications to predict choice of preference and provide significant recommendations. With the help of the advantage of deep learning in modeling different types of data and due to the dynamic change of user preference, building a deep model can better understand users demand and further improve quality of recommendation. In this paper, deep neural network models for recommender system are evaluated. Most of deep neural network models in recommender system focus on the classical collaborative filtering user-item setting. Deep learning models demonstrated high level features of complex data can be learned instead of using metadata which can significantly improve accuracy of recommendation. Even though deep learning poses a great impact in various areas, applying the model to a recommender system have not been fully exploited and still a lot of improvements can be done both in collaborative and content-based approach while considering different contextual factors.

Keywords: big data, decision making, deep learning, recommender system

Procedia PDF Downloads 483

Authors: Amir Deljoo

Abstract:

The endeavor to comprehend the existence have been the center of thought for human in form of different disciplines and now basically in physics as the theory of everything. Here, after a brief review of the basic frameworks of thought, and a history of thought since ancient up to present, a logical methodology is presented based on a core axiom after which a function, a proto-field and then a coordinates are explained. Afterwards a generalization to Standard Model is proposed as General Standard Model which is believed to be the base of the Unified Theory of Everything.

Keywords: general relativity, grand unified theory, quantum mechanics, standard model, theory of everything

Procedia PDF Downloads 105

Authors: Mohammad Javad Mollakazemi, Farhad Asadi, Aref Ghafouri

Abstract:

Performance of different filtering approaches depends on modeling of dynamical system and algorithm structure. For modeling and smoothing the data the evaluation of posterior distribution in different filtering approach should be chosen carefully. In this paper different filtering approaches like filter KALMAN, EKF, UKF, EKS and smoother RTS is simulated in some trajectory tracking of path and accuracy and limitation of these approaches are explained. Then probability of model with different filters is compered and finally the effect of the noise variance to estimation is described with simulations results.

Keywords: Gaussian approximation, Kalman smoother, parameter estimation, noise variance

Procedia PDF Downloads 445

Authors: Yehjune Heo

Abstract:

As biometric systems become widely deployed, the security of identification systems can be easily attacked by various spoof materials. This paper contributes to finding a reliable and practical anti-spoofing method using Convolutional Neural Networks (CNNs) based on the types of loss functions and optimizers. The types of CNNs used in this paper include AlexNet, VGGNet, and ResNet. By using various loss functions including Cross-Entropy, Center Loss, Cosine Proximity, and Hinge Loss, and various loss optimizers which include Adam, SGD, RMSProp, Adadelta, Adagrad, and Nadam, we obtained significant performance changes. We realize that choosing the correct loss function for each model is crucial since different loss functions lead to different errors on the same evaluation. By using a subset of the Livdet 2017 database, we validate our approach to compare the generalization power. It is important to note that we use a subset of LiveDet and the database is the same across all training and testing for each model. This way, we can compare the performance, in terms of generalization, for the unseen data across all different models. The best CNN (AlexNet) with the appropriate loss function and optimizers result in more than 3% of performance gain over the other CNN models with the default loss function and optimizer. In addition to the highest generalization performance, this paper also contains the models with high accuracy associated with parameters and mean average error rates to find the model that consumes the least memory and computation time for training and testing. Although AlexNet has less complexity over other CNN models, it is proven to be very efficient. For practical anti-spoofing systems, the deployed version should use a small amount of memory and should run very fast with high anti-spoofing performance. For our deployed version on smartphones, additional processing steps, such as quantization and pruning algorithms, have been applied in our final model.

Keywords: anti-spoofing, CNN, fingerprint recognition, loss function, optimizer

Procedia PDF Downloads 140

Authors: Mohamed Lotfy, Ghada Soliman

Abstract:

Recognizing Arabic dot-matrix digits is a challenging problem due to the unique characteristics of dot-matrix fonts, such as irregular dot spacing and varying dot sizes. This paper presents an approach for recognizing Arabic digits printed in dot matrix format. The proposed model is based on Convolutional Neural Networks (CNN) that take the dot matrix as input and generate embeddings that are rounded to generate binary representations of the digits. The binary embeddings are then used to perform Optical Character Recognition (OCR) on the digit images. To overcome the challenge of the limited availability of dotted Arabic expiration date images, we developed a True Type Font (TTF) for generating synthetic images of Arabic dot-matrix characters. The model was trained on a synthetic dataset of 3287 images and 658 synthetic images for testing, representing realistic expiration dates from 2019 to 2027 in the format of yyyy/mm/dd. Our model achieved an accuracy of 98.94% on the expiry date recognition with Arabic dot matrix format using fewer parameters and less computational resources than traditional CNN-based models. By investigating and presenting our findings comprehensively, we aim to contribute substantially to the field of OCR and pave the way for advancements in Arabic dot-matrix character recognition. Our proposed approach is not limited to Arabic dot matrix digit recognition but can also be extended to text recognition tasks, such as text classification and sentiment analysis.

Keywords: computer vision, pattern recognition, optical character recognition, deep learning

Procedia PDF Downloads 99

Authors: Jihan Abdulazeez Ahmed, Adnan Mohsin Abdulazeez Brifcani

Abstract:

This paper suggests a new internal architecture of holon based on feature selection model using the combination of Bees Algorithm (BA) and Artificial Neural Network (ANN). BA is used to generate features while ANN is used as a classifier to evaluate the produced features. Proposed system is applied on the Wine data set, the statistical result proves that the proposed system is effective and has the ability to choose informative features with high accuracy.

Keywords: artificial neural network, bees algorithm, feature selection, Holon

Procedia PDF Downloads 458

Authors: S. Ghafoori, A. Al-Harbi, B. Al-Ajmi, A. Al-Shaalan, A. Al-Ajmi, M. Ali Juma

Abstract:

Seismic methods play an important role in the exploration for hydrocarbon reservoirs. However, the success of the method depends strongly on the reliability of the measured or predicted information regarding the velocity of sound in the media. Speed of sound has been used to study the thermodynamic properties of fluids. In this study, experimental data are reported and analyzed on the speed of sound in toluene and octane binary mixture. Three-factor three-level Box-Benhkam design is used to determine the significance of each factor, the synergetic effects of the factors, and the most significant factors on speed of sound. The developed mathematical model and statistical analysis provided a critical analysis of the simultaneous interactive effects of the independent variables indicating that the developed quadratic models were highly accurate and predictive.

Keywords: experimental design, octane, speed of sound, toluene

Procedia PDF Downloads 278

Authors: Prateeksha Mahamallik, Anjali Pal

Abstract:

In the present study, Co(II)-adsolubilized surfactant modified alumina (SMA) was prepared, and methylene blue (MB) degradation was carried out on Co-SMA surface by visible light photo-Fenton process. The entire reaction proceeded on solid surface as MB was embedded on Co-SMA surface. The reaction followed zero order kinetics. Response surface methodology (RSM) and artificial neural network (ANN) were used for modeling the decolorization of MB by photo-Fenton process as a function of dose of Co-SMA (10, 20 and 30 g/L), initial concentration of MB (10, 20 and 30 mg/L), concentration of H2O2 (174.4, 348.8 and 523.2 mM) and reaction time (30, 45 and 60 min). The prediction capabilities of both the methodologies (RSM and ANN) were compared on the basis of correlation coefficient (R2), root mean square error (RMSE), standard error of prediction (SEP), relative percent deviation (RPD). Due to lower value of RMSE (1.27), SEP (2.06) and RPD (1.17) and higher value of R2 (0.9966), ANN was proved to be more accurate than RSM in order to predict decolorization efficiency.

Keywords: adsolubilization, artificial neural network, methylene blue, photo-fenton process, response surface methodology

Procedia PDF Downloads 260