Dropout Ratio Research Articles

The Flemish Interactive Coaching Monitoring System

As large-scale coach education programs receive a growing amount of attention and investment (e.g., human and financial resources), the case for increased understanding of their impact is a pressing matter. In this paper, the authors outlined the creation of the Flemish Interactive Coaching Monitoring System (FICOMS) within the Flemish School for Coach Education (Belgium). The FICOMS is a data warehouse consisting of multiple databases, which was set up in 2019 to integrate data on coach education and coach certifications (1960–present), active coaches within club-organized sports (2014–present) and sport clubs, sports participants, and sports infrastructure. The FICOMS provides a variety of interactive and externally facing dashboards with useful statistics on coach education and coaching in Flanders. For example, the evolution of dropout ratios of qualified versus nonqualified coaches in sports clubs and sports federations can be identified, as well as the evolution of the percentage of qualified coaches in a specific sport, sports federation, or gender, or regional differences. By describing the main characteristics of FICOMS and sharing some emerging insights and early possibilities, the authors aimed to clarify the potential of this information technology for different stakeholders, such as governments, policymakers, sports federations, Olympic committees, education partners, municipalities, and researchers.

Vaccination coverage and adherence to a dengue vaccination program in the state of Paraná, Brazil

The success of vaccination programs depends on the level of acceptance of the vaccine to achieve high vaccine coverage rates (VCR). Vaccine hesitancy is a challenge, especially concerning new vaccines. Dengue vaccine, Dengvaxia®, was licensed in Brazil in 2015 and implemented, in a pioneering publicly-funded initiative in the state of Paraná, between 2016 and 2018. The vaccination program took place in five phases in the 30 municipalities most affected by dengue in the state, targeting individuals from nine to 44 years-old in two cities and from 15 to 27 years-old in the other 28 municipalities, totaling a target population of 500,000 individuals. A cross-sectional descriptive study was carried out to assess VCR and adherence to the dengue vaccine in this program. VCR, dropout ratio (DR), and compliance with the vaccination schedule (CVS) were analyzed by sex, age group, and municipality size. A total of 302,603 individuals (60.5%) received ≥ 1 dose, 44.2% received ≥ 2 doses, and 28.6% 3 doses. The DR was 52.8%. Among individuals who started vaccination, 40.6% achieved CVS. The highest VCR, highest CVS, and lowest DR occurred in the age group from 9 to 14 years old and from 28 to 44 years old and in smaller municipalities. A greater proportion of men started vaccination (male 64.0%; female 57.1%) however, the DR was higher in men (male 55.4%; female 49.9%), and a higher percentage of women completed the vaccination schedule according to the recommendations (CVS male 37.8%; female 43.6%). Differences were noted in the CVS according to the initial phase of the program (first phase 50.8%; second phase 18.8%). The heterogeneity in vaccine uptake and compliance according to sex, age, and municipality size suggests the need for differentiated strategies to address challenges with new and multiple-dose vaccines.

Challenges to Learners with Disabilities in the Higher Education Institutions in Pakistan: A Review

Due to numerous reasons, educating learners with disabilities is an uphill task in Pakistan. Low enrolment and high dropout ratio of such learners need the attention of all stakeholders. The study inquired into the problems faced by learners with disabilities in Universities of Pakistan. The study, by reviewing the literature, examines the accessibility of learners with disabilities in universities, teachers' training, teaching methods, attitudinal challenges, and support services. The study concludes that these students face attitudinal problems both from fellow students and teachers; lack of access to classrooms, libraries, administrative offices; unfriendly teaching methods; untrained teaching staff; hostile environment; lack of assistive devices and transport to, from and inside the campus. The study recommends accessible buildings, accessible transport, capacity development of faculty in special and inclusive education, access to library and administration, curriculum revision, and implementation of policies regarding the inclusive education system.

Day-ahead spatiotemporal wind speed forecasting using robust design-based deep learning neural network

The operations of power systems are becoming increasingly challenging due to the high penetration of renewable power generation, which is uncertain and stochastic. Highly accurate day-ahead renewable power forecasting helps operators in a dispatch center schedule power generations for conventional thermal generators. Efficient day-ahead renewable power forecasting requires information from widely separated locations. This paper proposes a novel day-ahead spatiotemporal wind speed forecasting method based on a convolutional neural network (CNN), which is an image-based deep learning method. This paper exploits a robust design based on Taguchi’s orthogonal array to determine the numbers of cascade/parallel layers, convolutions in each layer, kernels and hidden layers, as well as the kernel size, zero-padding and dropout ratio in the proposed CNN to ensure that the proposed CNN is insensitive to the seasonal characteristics of wind speeds while retaining high accuracy. In this paper, the day-ahead forecasting of wind speed at an offshore wind farm (Fuhai) near Taiwan is performed using historical wind speeds at seven sites in Taiwan, China, South Korea and the Philippines. Simulation results reveal that the proposed robust design-based CNN outperforms existing methods.

Element extraction and convolutional neural network-based classification for blue calico

Blue calico is a highly valued folk handicraft that forms part of China’s national intangible cultural heritage. Thus, blue calico is a worthy target for reconstruction using modern image processing technology. Extracting the visual components or elements of a blue calico pattern is one way to capture the underlying design and enable innovation in traditional patterns using modern techniques. This paper presents a method of element extraction and classification based on a smart convolutional neural network (CNN), with an improved CifarNet structure, which we call CalicoNet. Initially, the algorithm for element extraction is implemented to generate element samples of blue calico. This process includes gray scaling, binarization, and contour extraction. We construct a data set of elements with 12 types. Then, four critical hyper-parameters, the batch-size, dropout ratio, learning rate, and pooling strategy, are optimized by a comparative analysis. A combination classifier strategy is subsequently added to the fully connected layers to strengthen the feature expression in the corresponding classes. Finally, the superiority of the proposed CalicoNet is verified through a comparison with other sophisticated CNNs. Experimental results demonstrate that CalicoNet achieves a validation accuracy of 99.2% for the training set, a total time of 1.13 hours for the whole data set, and a test mean accuracy precision of 98.66%. The robust performance of the proposed method across the element data set indicates that CalicoNet is a promising approach for element extraction and classification.

Dry Eye and Quantitative and Qualitative Changes of Meibomian Glands in Patients With Pemphigus.

To evaluate dry eye and quantitative and qualitative changes of meibomian glands with meibography in patients with pemphigus. Twenty left eyes of 20 patients with pemphigus (group 1) and 20 left eyes of 20 age- and sex-matched healthy volunteers (group 2) were enrolled. Dry eye was evaluated with Oxford scale scoring, tear film breakup time, Schirmer 1 tests, and Ocular Surface Disease Index score assessments. Morphological changes and dropout ratio of meibomian glands were evaluated by upper and lower eyelid meibography (SL-D701; Topcon, IJssel, The Netherlands). Meibomian gland dropout ratio was scored for each eyelid from grade 0 (no loss) through grade 3 (lost area >2/3 of the total meibomian gland area). The mean ages of group 1 and group 2 were 51.8 ± 11.1 (range: 32-73) and 50.4 ± 9.6 years (range: 32-70), respectively (P = 0.672). Schirmer 1 and tear film breakup time values were lower and Oxford and Ocular Surface Disease Index scores were higher in group 1 when compared with group 2, and the differences were statistically significant (P = 0.01, P < 0.001, P = 0.001, and P < 0.001, respectively). Upper eyelid, lower eyelid, and total eyelid meiboscore values were significantly higher in group 1 (P = 0.005, P = 0.018, and P = 0.002, respectively). Morphological changes in meibomian glands were detected in 16 eyes (80%) among group 1. Between patients with pemphigus and those who were healthy, there were significant differences in the results of dry eye tests and meibomian gland morphology. Patients with pemphigus should be monitored for dry eye and meibomian gland dysfunction and be promptly treated to prevent the serious consequences of dry eye.

Research on pig face recognition model based on keras convolutional neural network

Traditionally, the RFID technology is generally used for the identification of pigs in vivo. However, the method of electronic ear tags and ear tags will cause great pain to the pigs, then ear tags will easily fall off during the pig’s activities, increasing the operating cost of the enterprise. This paper uses the powerful feature learning and feature expression capabilities of convolutional neural networks in deep learning to automatically learn the facial features of pigs. Use the Image Data Generator that comes with Keras to perform data enhancement on the pig face pictures of ten pigs and generate pig face dataset. This paper proposes a convolutional neural network model based on LeNet-5 for facial image recognition of pigs. Experimental comparisons were performed by using SGD, Adam and rmsprop optimizers with dropout ratios of 0.3, 0.5and 0.7. Experiments show that when the SGD optimizer is used and dropout is 0.3, the model recognition rate is the highest, which can reach 97.6%.

Relationship between Crime and Economic Conditions in Sindh: A Time Series Approach from 1984-2015

Using the time series data from 1984 to 2015, this study attempts to explore Sindh economic situation and the relationship between criminal activities. Three Variables are used for economic conditions, such as crime rate, dropout ratio and unemployment. We check their relationship with the reported crime. Enhanced Dicky Fuller test for unit root process indicates that all variables are stationary at the first level. For long-term relationships, Johanson-Cointegration technology has been applied. The results of the statistical process show that dropout ratio and unemployment are closely related to crime.VCM has been applied to check the short-run relationship between the variables. VCM results suggested that the model we estimate is divergent. Divergent model mean that there is no adjustment from long-run to short-run between variables as they are going away, if we increase the lag length, the model can become divergent but due to crime data unavailability it was difficult to increase the observations and the lags as well. Study gives evidence that economic conditions have significant impact on crimes and increasing dropout which is Positive related with crime in Sindh. It is also shown that the crime is influenced by economic condition. Government is capable to reduce that threat through effective target policies and legislation. The empirical results of this study will enhance understanding of the role of public sector policy formation in promoting national productive capacity by uplifting the positive effect of the Sindh economy.

Forecasting and Avoiding Student Dropout Using the K-Nearest Neighbor Approach

All India School Education Survey of 2018–2019 has put a total of 13,06,992 schools and enrollment figure as 22,67,19,283 number of students in these schools in India, and the problem is continuing studies at school or college for many students is difficult due to reasons such as financial problem, domestic problems at home, being unable to cope up with studies, language or medium of education and many other problems like this. The dropout ratio in 2018–2019 for secondary level classes IX and X is at 17% as compared to 4% in primary classes I to V, and to minimize the dropout ratio, the technique of prediction is proposed; in the field of machine learning and data mining, the most widely used prediction method is K-nearest neighbors. This method is more versatile and simple and can handle different types of data in the process of prediction. The prediction of students is classified into dropout or no-dropout category and hence enables the teacher to counsel the students who are at risk of dropping out.

Student Dropout Prediction &amp; Educational Data Mining

Educational data like students performance is very important to study and analyze and to improve the quality of education. The study concerned to data mining techniques with educational data is known as Educational Data Mining (EDM). This study finds knowledge and interesting patterns in educational organization. Students performance are the subject mainly concerned to find the qualitative model based on student’s personal and social factors then classify and predict the student performance. Proper counseling to underperforming students can reduce dropout ratio and help them to continue their studies.

Multi-Factor Operating Condition Recognition Using 1D Convolutional Long Short-Term Network.

For a diesel engine, operating conditions have extreme importance in fault detection and diagnosis. Limited to various special circumstances, the multi-factor operating conditions of a diesel engine are difficult to measure, and the demand of automatic condition recognition based on vibration signals is urgent. In this paper, multi-factor operating condition recognition using a one-dimensional (1D) convolutional long short-term network (1D-CLSTM) is proposed. Firstly, a deep neural network framework is proposed based on a 1D convolutional neural network (CNN) and long short-Term network (LSTM). According to the characteristics of vibration signals of a diesel engine, batch normalization is introduced to regulate the input of each convolutional layer by fixing the mean value and variance. Subsequently, adaptive dropout is proposed to improve the model sparsity and prevent overfitting in model training. Moreover, the vibration signals measured under 12 operating conditions were used to verify the performance of the trained 1D-CLSTM classifier. Lastly, the vibration signals measured from another kind of diesel engine were applied to verify the generalizability of the proposed approach. Experimental results show that the proposed method is an effective approach for multi-factor operating condition recognition. In addition, the adaptive dropout can achieve better training performance than the constant dropout ratio. Compared with some state-of-the-art methods, the trained 1D-CLSTM classifier can predict new data with higher generalization accuracy.

Therapeutic Dose of Amitriptyline for Older Patients with Burning Mouth Syndrome.

ObjectiveTo assess the therapeutic dose and safety of amitriptyline and the outcome following treatment with amitriptyline among older patients with burning mouth syndrome (BMS).Methods187 consecutive patients were prescribed amitriptyline as a first-line medication from April 2016 to September 2018 and followed-up for >1 month. Patients were divided into 3 groups: group 1, 113 patients aged <65 years; group 2, 52 patients aged between 65 and 74 years; and group 3, 22 patients aged 75 years or older. The visual analog scale (VAS), Pain Catastrophizing Scale (PCS), Somatic Symptom Scale-8 (SSS-8), Patient Global Impression of Change (PGIC), and Short-form McGill Pain Questionnaire (SF-MPQ) were used for analysis.ResultsThirty-two patients (17 in group 1, 10 in group 2, and 5 in group 3) stopped taking amitriptyline due to side effects. There were no differences among the groups with respect to sex; scores of VAS, PCS, and SSS-8; and drop-out ratio. There were no significant differences in the VAS, PCS, and PGIC scores among the groups after 1 month. The mean daily dose after 1 month was 20.4 ± 8.6 mg in group 1, 17.3 ± 8.7 mg in group 2, and 13.2 ± 5.8 mg in group 3; this difference was significant (p value = 0.003). About 76% of patients showed improvements in their symptoms (PGIC ≥ 3). About 90% of patients reported side effects. No serious side effects occurred.ConclusionThe therapeutic dose of amitriptyline may be lower for older BMS patients than for younger patients.

The patterns of the dropout of doctoral students

Compared to the trends of the OECD countries, the proportion of graduated higher educational students is low in Hungary, especially in the doctoral education; however, the need for doctoral education has been increased due to the expansion in the past two and half decades. According to the report of the European Committee, the ratio of those having a PhD degree is 0.8 per mille among the 25- to 34-year-old Hungarian population that is not advantageous. As a causal factor, the dropout can be detected. The aim of my investigation is to detect the pattern of the dropout of the doctoral students. Who are those who quit their studies before finishing it? In which field is the ratio of graduation the lowest? I hypothesize that approximately one fifth of the students enrolled for a doctoral course quit their studies during the education. According to the educational fields, the dropout ratio is the lowest in the natural, medical, and agricultural sciences, whereas it is the highest in the arts, human, and social sciences. The basis of the current research is provided by a huge national database entitled Higher Educational Informational System, containing the data of doctoral students enrolled in the autumn semester of the school years 2010/2011 and 2014/2015.

Practical or theoretical persistence?: The investigation of (f)actors influencing students’ persistence at three levels

Higher educational dropout is a major education policy issue that can be influenced by several factors. In addition to the family background, it is necessary to mention the motivation for further education as an individual factor which has a complex effect. Another possible individual cause can be the attractiveness of the labor market. Due to the ratio of students dropping out of higher education in Hungary, it can be suspected that students’ intensive work contributes to weaker learning outcomes, resulting dropout finally. In this context, however, the decisive role of the different work values and working attitudes is also unquestionable. Other institutional factors such as the country of the institution or the type of financing of the training cannot be ignored as well. Accordingly, in our research, we investigated individual, institutional, and sociodemographic factors affecting persistence through the TESCEE 2015 (N = 2015) database. Factors influencing persistence were measured by linear regression analysis with the application of two-sample t-test to measure the between-group differences. Regarding socio-demographic factors, the father’s educational level showed a significant impact on a negative while the mother’s employment in a positive way, furthermore, gender presented a trend effect. Institutional factors by themselves are not remarkable; however, some individual factor can increase their impact. At the individual level, the significant effect of career office membership and work values could be detected. Our results can contribute to the recognition of the relationships behind the high ratio of dropout and the identification of factors that can promote persistence, which can support to reduce the dropout ratio at a national and international level.

Soft-shell Shrimp Recognition Based on an Improved AlexNet for Quality Evaluations

Shrimp quality evaluations fulfill an essential role in producing high-value shrimp products. The presence of soft-shell shrimp deteriorates the quality of shrimp products. The biggest challenge in preventing this is the similarity in appearance of soft-shell (s-shrimp) and sound (o-shrimp) shrimp from an imaging perspective. This similarity imposes significant limitations on distinguishing them with traditional machine vision methods. To circumvent this problem, a novel method based on deep convolutional neural networks (Deep-ShrimpNet) is proposed. Initially, several image processing steps were performed to normalize the shrimp image. Furthermore, four critical hyper-parameters (i.e., batch-size, dropout ratio, learning rate and number (size) of local receptive fields) were optimized by a comparative analysis. Additionally, the self-learned combined features in each convolutional layer were visualized to explore the internal mechanism of Deep-ShrimpNet. To obtain the efficient strategy, an ablation study was also performed by removing layers of the CNNs. Finally, the superiority of the proposed algorithm was verified through a comparison with other sophisticated CNNs. In a test dataset, Deep-ShrimpNet achieved a mean accuracy precision (mAP) of 0.972 and modeling time of 0.54 h. The robust performance of the proposed method across the shrimp dataset indicates that Deep-ShrimpNet is promising for on-line shrimp classification and quality measurement.

Deep Learning for Emotion Recognition in Affective Virtual Reality and Music Applications

This paper presents a deep learning approach to emotion recognition as applied to virtual reality and music predictive analytics. Firstly, it investigates the deep parameter tuning of the multi-hidden layer neural networks, which are also commonly referred to simply as deep networks that are used to conduct emotion detection in virtual reality (VR)- electroencephalography (EEG) predictive analytics. Deep networks have been studied extensively over the last decade and have shown to be among the most accurate methods for predictive analytics in image recognition and speech processing domains. However, most predictive analytics deep network studies focus on the shallow parameter tuning when attempting to boost prediction accuracies, which includes deep network tuning parameters such as number of hidden layers, number of hidden nodes per hidden layer and the types of activation functions used in the hidden nodes. Much less effort has been put into investigating the tuning of deep parameters such as input dropout ratios, L1 (lasso) regularization and L2 (ridge regularization) parameters of the deep networks. As such, the goal of this study is to perform a parameter tuning investigation on these deep parameters of the deep networks for predicting emotions in a virtual reality environment using electroencephalography (EEG) signal obtained when the user is exposed to immersive content. The results show that deep tuning of deep networks in VR-EEG can improve the accuracies of predicting emotions. The best emotion prediction accuracy was improved to over 96% after deep tuning was conducted on the deep network parameters of input dropout ratio, L1 and L2 regularization parameters. Secondly, it investigates a similar possible approach when applied to 4-quadrant music emotion recognition. Recent studies have been characterizing music based on music genres and various classification techniques have been used to achieve the best accuracy rate. Several researches on deep learning have shown outstanding results in relation to dimensional music emotion recognition. Yet, there is no concrete and concise description to express music. In regards to this research gap, a research using more detailed metadata on twodimensional emotion annotations based on the Russell’s model is conducted. Rather than applying music genres or lyrics into machine learning algorithm to MER, higher representation of music information, acoustic features are used. In conjunction with the four classes classification problem, an available dataset named AMG1608 is feed into a training model built from deep neural network. The dataset is first preprocessed to get full access of variables before any machine learning is done. The classification rate is then collected by running the scripts in R environment. The preliminary result showed a classification rate of 46.0%.

A Comparison Between Major Artificial Intelligence Models for Crop Yield Prediction: Case Study of the Midwestern United States, 2006–2015

This paper compares different artificial intelligence (AI) models in order to develop the best crop yield prediction model for the Midwestern United States (US). Through experiments to examine the effects of phenology using three different periods, we selected the July–August (JA) database as the best months to predict corn and soybean yields. Six different AI models for crop yield prediction are tested in this research. Then, a comprehensive and objective comparison is conducted between the AI models. Particularly for the deep neural network (DNN) model, we performed an optimization process to ensure the best configurations for the layer structure, cost function, optimizer, activation function, and drop-out ratio. In terms of mean absolute error (MAE), our DNN model with the JA database was approximately 21–33% and 17–22% more accurate for corn and soybean yields, respectively, than the other five AI models. This indicates that corn and soybean yields for a given year can be forecasted in advance, at the beginning of September, approximately a month or more ahead of harvesting time. A combination of the optimized DNN model and spatial statistical methods should be investigated in future work, to mitigate partly clustered errors in some regions.

Adaptive gamification in e-learning based on students’ learning styles

ABSTRACT With the removal of the barriers of time and distance, E-learning platforms have attracted millions of learners, but these platforms are experiencing a significant drop-out ratio. One of the primary reasons for this problem is the lack of motivation among the learners because of the similar learning experience provided to them despite their varying learning styles. Different researchers have introduced gamification as a solution for students’ engagement. The technique has improved engagement levels a bit, but it is not as useful as it was expected to be. One of the primary problems with gamification elements is their inability to induce intrinsic motivation among learners. To target this issue, we have proposed a framework that identifies the learning style of students based on their interactions with the system and provides an adaptive gamification experience according to their identified learning dimensions. The results of the experiments show that the motivation of learners increases by 25%, and the drop-out ratio is reduced by 26%.

Integrated development of students: The telangana scheduled castes development department

This research paper explores the programs and schemes for integrated and all-round development of students in Scheduled Castes in Telangana. The Scheduled Caste Development Department is striving for the development of scheduled castes. The scheduled castes population is 5.4 million, which accounts for 15.43% of the total population. According to the census (2011) report, the literacy rate is 66.46%. The literacy level is low among Scheduled Castes with 52.52%, and the drop-out ratio is high in Scheduled Castes compared to that of all communities. Many studies revealed a high rate of wastage, stagnation, and inequality of the opportunity. The Indian Constitution states that India as a welfare state. With the salient features in the Preamble and the Directive Principles, The SC Development Department dedicated to the integrated development of the scheduled castes. The Department's primary objectives are educational advancement, socio-economic

Formación como elemento de cohesión social en Europa. Análisis comparativo de sinergias y recursos ejecutados, empleando una muestra de medidas operativas (programas) en el marco de las políticas europeas de promoción de la formación profesional y vocacional (VET), lucha contra el abandono formativo y académico temprano (ELET) y recualificación

The implementation of the commitments agreed upon in the Europe 2020 Strategy, coupled in a transversally way, the integrated development of multiple policies and work strategies in the educational, social and professional areas. In particular, and by elocution, the commitments regarding the attainment of a modernized vocational training, aimed at labour integration, the reduction of the education and training drop-out ratios, and in turn, the integration through training, of cultural or socio-economically disadvantaged groups. Under this context, the European Council, using the different tools available for the implementation of active policies, put into operation different measures in place to implement and complete the agreed commitments. Within these measures, programmes of diverse nature were applied that in some way incorporated, potended and contributed to the achievement of the commitments. In order to know the synergies in the interventions and actions carried out, a sample was formed from different programmes implemented during the last years, using different resources, that allowed to establish the orientation of the executed actions and the degree of weight and representation of the different variables that intervene in the analysis. The comparative analysis of the proposed indicators allowed to establish, in addition, the classification of the interventions and thus be able to know a defined sample. The main results obtained are the weight of Vocational Education anda Training and Professional Guidance as the backbone of the programmes analyzed, among other conclusions and results.