Data scarcity is a major challenge when training deep learning (DL) models. DL demands a large amount of data to achieve exceptional performance. Unfortunately, many applications have small or inadequate data to train DL frameworks. Usually, manual labeling is needed to provide labeled data, which typically involves human annotators with a vast background of knowledge. This annotation process is costly, time-consuming, and error-prone. Usually, every DL framework is fed by a significant amount of labeled data to automatically learn representations. Ultimately, a larger amount of data would generate a better DL model and its performance is also application dependent. This issue is the main barrier for many applications dismissing the use of DL. Having sufficient data is the first step toward any successful and trustworthy DL application. This paper presents a holistic survey on state-of-the-art techniques to deal with training DL models to overcome three challenges including small, imbalanced datasets, and lack of generalization. This survey starts by listing the learning techniques. Next, the types of DL architectures are introduced. After that, state-of-the-art solutions to address the issue of lack of training data are listed, such as Transfer Learning (TL), Self-Supervised Learning (SSL), Generative Adversarial Networks (GANs), Model Architecture (MA), Physics-Informed Neural Network (PINN), and Deep Synthetic Minority Oversampling Technique (DeepSMOTE). Then, these solutions were followed by some related tips about data acquisition needed prior to training purposes, as well as recommendations for ensuring the trustworthiness of the training dataset. The survey ends with a list of applications that suffer from data scarcity, several alternatives are proposed in order to generate more data in each application including Electromagnetic Imaging (EMI), Civil Structural Health Monitoring, Medical imaging, Meteorology, Wireless Communications, Fluid Mechanics, Microelectromechanical system, and Cybersecurity. To the best of the authors’ knowledge, this is the first review that offers a comprehensive overview on strategies to tackle data scarcity in DL.
This paper deals to how to estimate points non measured spatial data when the number of its terms (sample spatial) a few, that are not preferred for the estimation process, because we also know that whenever if the data is large, the estimation results of the points non measured to be better and thus the variance estimate less, so the idea of this paper is how to take advantage of the data other secondary (auxiliary), which have a strong correlation with the primary data (basic) to be estimated single points of non-measured, as well as measuring the variance estimate, has been the use of technique Co-kriging in this field to build predictions spatial estimation process, and then we applied this idea to real data in th
... Show MoreAim: This study aimed to assessing orthodontic knowledge and attitude among general dentists and non-orthodontic specialists. Background: Early detection of orthodontic disorders is essentialin motivating patients to intervene prior to long term complications when the disorders are not recongised. Methods: A questionnaire was distributed amongst dentistsother than orthodontists. This questionnaire consisted of three sections. The first one aimed to collect demographic, educational level and practice type information. Further two sections consisted of closed-end questions designed to evaluateknowledge and attitude of orthodontics. Results: A total of 313 responses to the survey were submitted. No significant correlation was observed, e
... Show MoreOur aim was to investigate the inclusion of sexual and reproductive health and rights (SRHR) topics in medical curricula and the perceived need for, feasibility of, and barriers to teaching SRHR. We distributed a survey with questions on SRHR content, and factors regulating SRHR content, to medical universities worldwide using chain referral. Associations between high SRHR content and independent variables were analyzed using unconditional linear regression or χ2 test. Text data were analyzed by thematic analysis. We collected data from 219 respondents, 143 universities and 54 countries. Clinical SRHR topics such as safe pregnancy and childbirth (95.7%) and contraceptive methods
Abstract
The current research aims to identify the level of E-learning among middle school students, the level of academic passion among middle school students, and the correlation between e-learning and academic passion among middle school students. In order to achieve the objectives of the research, the researcher developed two questionnaires to measure the variables of the study (e-learning and study passion) among students, these two tools were applied to the research sample, which was (380) male and female students in the first and second intermediate classes. The research concluded that there is a relationship between e-learning and academic passion among students.
The emergence of mixed matrix membranes (MMMs) or nanocomposite membranes embedded with inorganic nanoparticles (NPs) has opened up a possibility for developing different polymeric membranes with improved physicochemical properties, mechanical properties and performance for resolving environmental and energy-effective water purification. This paper presents an overview of the effects of different hydrophilic nanomaterials, including mineral nanomaterials (e.g., silicon dioxide (SiO2) and zeolite), metals oxide (e.g., copper oxide (CuO), zirconium dioxide (ZrO2), zinc oxide (ZnO), antimony tin oxide (ATO), iron (III) oxide (Fe2O3) and tungsten oxide (WOX)), two-dimensional transition (e.g., MXene), metal–organic framework (MOFs), c
... Show MoreThe basic objective of the research is to study the quality of the water flow service in the Directorate of Karbala sewage and how to improve it after identifying the deviations of the processes and the final product and then providing the possible solutions in addressing the causes of the deviations and the associated quality gaps. A number of quality tools were used and applied to all data Stations with areas and activities related to the drainage of rainwater, as the research community determines the stations of lifting rainwater in the Directorate of the streams of Karbala holy, and the station was chosen Western station to apply the non-random sampling method intended after meeting a number of. It is one of the largest and m
... Show MoreThis study reports testing results of the transient response of T-shape concrete deep beams with large openings due to impact loading. Seven concrete deep beams with openings including two ordinary reinforced, four partially prestressed, and one solid ordinary reinforced as a reference beam were fabricated and tested. The effects of prestressing strand position and the intensity of the impact force were investigated. Two values for the opening’s depth relative to the beam cross-section dimensions were inspected under the effect of an impacting mass repeatedly dropped from different heights. The study revealed that the beam’s transient deflection was increased by about 50% with gre
This paper deals with an analytical study of the flow of an incompressible generalized Burgers’ fluid (GBF) in an annular pipe. We discussed in this problem the flow induced by an impulsive pressure gradient and compare the results with flow due to a constant pressure gradient. Analytic solutions for velocity is earned by using discrete Laplace transform (DLT) of the sequential fractional derivatives (FD) and finite Hankel transform (FHT). The influences of different parameters are analyzed on a velocity distribution characteristics and a comparison between two cases is also presented, and discussed in details. Eventually, the figures are plotted to exhibit these effects.
Merging biometrics with cryptography has become more familiar and a great scientific field was born for researchers. Biometrics adds distinctive property to the security systems, due biometrics is unique and individual features for every person. In this study, a new method is presented for ciphering data based on fingerprint features. This research is done by addressing plaintext message based on positions of extracted minutiae from fingerprint into a generated random text file regardless the size of data. The proposed method can be explained in three scenarios. In the first scenario the message was used inside random text directly at positions of minutiae in the second scenario the message was encrypted with a choosen word before ciphering
... Show More