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 research seeks to shed light on what you add intangible assets of benefit to the company and this antagonize pause for consideration because it makes the company in a good competitive position stimulates the rest of the companies to acquire those assets.
That many companies have achieved competitive advantages in the market do not even achieved monopolies increased the value and reaped extraordinary profits as a result of those assets which requires the need to be measured to determine the extent to which contribution in the emergence of the value added to the value of the company on the one hand and to make the presentatio
... Show MoreAbstract:
Organizations need today to move towards strategic innovation, which means the analysis of positions, especially the challenges faced by the change in the external environment, which makes it imperative for the organization that you reconsider their strategies and orientations and operations, a so-called re-engineering to meet those challenges and pressures. Now this research dilemma intellectual two-dimensional, yet my account in not Take writings and researchers effect strategic innovation in re-engineering business processes, according to science and to inform the researcher, and after the application represented in the non-application of such resear
... Show MoreThis research aims to examine the effectiveness of a teaching strategy based on the cognitive model of Daniel in the development of achievement and the motivation of learning the school mathematics among the third intermediate grade students in the light of their study of "Systems of Linear Equations”. The research was conducted in the first semester (1439/1440AH), at Saeed Ibn Almosaieb Intermediate School, in Arar, Saudi Arabia. A quasi-experimental design has been used. In addition, a (pre & post) achievement test (20 Questions) and a (pre & post) scale of learning motivation to the school mathematics (25 Items) have been applied on two groups: a control group (31Students), and an experimental group (29 Students). The resear
... Show MoreAbstract. Hassan FM, Mahdi WM, Al-Haideri HH, Kamil DW. 2022. Identification of new species record of Cyanophyceae in Diyala River, Iraq based on 16S rRNA sequence data. Biodiversitas 23: 5239-5246. The biodiversity and water quality of the Diyala River require screening water in terms of biological contamination, because it is the only water source in Diyala City and is used for many purposes. This study aimed to identify a new species record of Cynaophyceae and emphasize the importance of using molecular methods beside classic morphological approaches, particularly in the water-shrinkage-aqua system. Five different sites along Diyala River were selected for Cyanophyceae identification. Morphological examination and 16S rRNA sequen
... Show MoreThe Atmospheric Infrared Sounder (AIRS) on EOS/Aqua satellite provides diverse measurements of Methane (CH4) distribution at different pressure levels in the Earth's atmosphere. The focus of this research is to analyze the vertical variations of (CH4) volume mixing ratio (VMR) time-series data at four Standard pressure levels SPL (925, 850, 600, and 300 hPa) in the troposphere above six cities in Iraq from January 2003 to September 2016. The analysis results of monthly average CH4VMR time-series data show a significant increase between 2003 and 2016, especially from 2009 to 2016; the minimum values of CH4 were in 2003 while the maximum values were in 2016. The vertical distribution of CH4<
... Show MoreWith the development of cloud computing during the latest years, data center networks have become a great topic in both industrial and academic societies. Nevertheless, traditional methods based on manual and hardware devices are burdensome, expensive, and cannot completely utilize the ability of physical network infrastructure. Thus, Software-Defined Networking (SDN) has been hyped as one of the best encouraging solutions for future Internet performance. SDN notable by two features; the separation of control plane from the data plane, and providing the network development by programmable capabilities instead of hardware solutions. Current paper introduces an SDN-based optimized Resch