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.
The exchanges in various fields,like economics, science, culture, etc., have been enhanced unceasingly among different countries around the world in the twenty-first century, thus, the university graduate who masters one foreign language does not meet the need of the labor market in most countries.So, many universities began to develop new programs to cultivate students who can use more foreign languages to serve the intercultural communication. At the same time, there is more scientific research emerged which is related to the relationship between the second and third languages. This humble research seeks to explain the relevant concepts and analyze the real data collected from Shanghai International Studies University in China, to expl
... Show MoreAbstract
The present paper attempts to detect the level of (COVID-19) pandemic panic attacks among university students, according to gender and stage variables.
To achieve this objective, the present paper adopts the scale set up by (Fathallah et al., 2021), which has been applied electronically to a previous cross-cultural sample consisting of (2285) participants from Arab countries, including Iraq. The scale includes, in its final form, (69) optional items distributed on (6) dimensions: physical symptoms (13) items, psychological and emotional symptoms (12) items, cognitive and mental symptoms (11) items, social symptoms (8) items, general symptoms (13) items and daily living practices (12) items
... Show MoreIn Automatic Speech Recognition (ASR) the non-linear data projection provided by a one hidden layer Multilayer Perceptron (MLP), trained to recognize phonemes, and has previous experiments to provide feature enhancement substantially increased ASR performance, especially in noise. Previous attempts to apply an analogous approach to speaker identification have not succeeded in improving performance, except by combining MLP processed features with other features. We present test results for the TIMIT database which show that the advantage of MLP preprocessing for open set speaker identification increases with the number of speakers used to train the MLP and that improved identification is obtained as this number increases beyond sixty.
... Show MoreThis paper proposes a new encryption method. It combines two cipher algorithms, i.e., DES and AES, to generate hybrid keys. This combination strengthens the proposed W-method by generating high randomized keys. Two points can represent the reliability of any encryption technique. Firstly, is the key generation; therefore, our approach merges 64 bits of DES with 64 bits of AES to produce 128 bits as a root key for all remaining keys that are 15. This complexity increases the level of the ciphering process. Moreover, it shifts the operation one bit only to the right. Secondly is the nature of the encryption process. It includes two keys and mixes one round of DES with one round of AES to reduce the performance time. The W-method deals with
... Show MoreBrachycerous Dipteran species on alfalfa plant Medicago sativa surveyed in several regions of Iraq from March to November 2012. The study was registered 14 species belonging to nine genera and four families. The results showed that Limnophra quaterna, Atherigona laevigata and Atherigona theodori as new records to Iraq and new pests of alfalfa.
Sports skills in some individual games require physical and motor qualities to facilitate the process of skill performance and also require the instructor or trainer to use more than one strategy, method and way to bring the performance to the level of mastery and avoid injury. The aim of the research is to know the effect of using special exercises using tools and their effect on teaching the skill of a front shoulder circle. The research hypothesis is that using special exercises with tools has a positive effect on teaching the skill of a front shoulder circle on the rings apparatus. Research method: - The researchers used the experimental method by designing two equal groups, the control and the experimental, to suit the research
... Show MoreAs technology advances and develops, the need for strong and simple authentication mechanisms that can help protect data intensifies. The contemporary approach to giving access control is through graphical passwords comprising images, patterns, or graphical items. The objective of this review was to determine the documented security risks that are related to the use of graphical passwords, together with the measures that have been taken to prevent them. The review was intended to present an extensive literature review of the subject matter on graphical password protection and to point toward potential future research directions. Many attacks, such as shoulder surfing attacks, SQL injection attacks, and spyware attacks, can easily ex
... Show More