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.
E-learning applications according to the levels of enlightenment (STEM Literacy) for physics teachers in the secondary stage. The sample consists of (400) teachers, at a rate of (200) males (50%), and (200)females (50%), distributed over (6) directorates of education in Baghdad governorate on both sides of Rusafa and Karkh. To verify the research goals, the researcher built a scale of e-learning applications according to the levels of STEM Literacy, which consists of (50) items distributed over (5) levels. The face validity of the scale and its stability were verified by extracting the stability coefficient through the internal consistency method “Alf-Cronbach”. The following statistical means were used: Pearson correlation coefficient,
... Show MoreTo maintain a sustained competitive position in the contemporary environment of knowledge economy, organizations as an open social systems must have an ability to learn and know how to adapt to rapid changes in a proper fashion so that organizational objectives will be achieved efficiently and effectively. A multilevel approach is adopted proposing that organizational learning suffers from the lack of interest about the strategic competitive performance of the organization. This remains implicit almost in all models of organizational learning and there is little focus on how learning organizations achieve sustainable competitive advantage . A dynamic model that captures t
... Show MoreThis research reviews the aesthetic variables that were founded according to (theatrical rehearsal) as one of the most important pillars on which the theatrical process is based, because of its necessity in developing theatrical art on several levels that helped the theatrical director in organizing his work, and this became clear through the research chapters represented in the first chapter (methodological framework) and the second chapter, which consisted of the first topic (the duality of watching / rehearsal) and the second topic (the applications of theatrical rehearsal in theatrical experiences), all the way to the third chapter (research procedures), which included the analysis of theatrical rehearsals (sharing on life), and the
... Show MoreParticulate matter (PM) emitted from diesel engine exhaust have been measured in terms of mass, using
99.98 % pure ethanol blended directly, without additives, with conventional diesel fuel (gas – oil),to
get 10 % , 15 %, 20 % ethanol emulsions . The resulting PM collected has been compared with those
from straight diesel. The engine used is a stationary single cylinder, variable compression ratio Ricardo
E6/US. This engine is fully instrumented and could run as a compression or spark ignition.
Observations showed that particulate matter (PM) emissions decrease with increasing oxygenate
content in the fuel, with some increase of fuel consumption, which is due to the lower heating value of
ethanol. The reduction in
The non static chain is always the problem of static analysis so that explained some of theoretical work, the properties of statistical regression analysis to lose when using strings in statistic and gives the slope of an imaginary relation under consideration. chain is not static can become static by adding variable time to the multivariate analysis the factors to remove the general trend as well as variable placebo seasons to remove the effect of seasonal .convert the data to form exponential or logarithmic , in addition to using the difference repeated d is said in this case it integrated class d. Where the research contained in the theoretical side in parts in the first part the research methodology ha
... Show MoreThe bandwidth requirements of telecommunication network users increased rapidly during the last decades. Optical access technologies must provide the bandwidth demand for each user. The passive optical access networks (PONs) support a maximum data rate of 100 Gbps by using the Orthogonal Frequency Division Multiplexing (OFDM) technique in the optical access network. In this paper, the optical broadband access networks with many techniques from Time Division Multiplexing Passive Optical Networks (TDM PON) to Orthogonal Frequency Division Multiplex Passive Optical Networks (OFDM PON) are presented. The architectures, advantages, disadvantages, and main parameters of these optical access networks are discussed and reported which have many ad
... Show More