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.
Improving" Jackknife Instrumental Variable Estimation method" using A class of immun algorithm with practical application
This article introduces a numerical study on heat exchange and corrosion coefficients of Zinc–water nanofluid stream in a circular tube fitted with swirl generator utilizing CFD emulation. Different forms of swirl generator which have the following properties of plain twisted tape (PTT) and baffle wings twisted tape (BTT) embeds with various ratio of twisting (y = 2.93, 3.91 and 4.89), baffle inclination angles (β = 0°, - 30° and 30) joined with 1%, 1.5% and 2% volume fraction of ZnO nanofluid were utilized for simulation. The results demonstrated that the heat and friction coefficients conducted by these two forms of vortex generator raised with Reynolds number, twist ratio and baffle inclination angles decreases. Likewise, t
... Show MoreThe study conducted to investigate the association between Helicobacter pylori infection and eye diseases (Glaucoma, Cataract, CSR and Uveitis). One hundred and four patients with multiple eye disorders (10-80) years were observed from 10/9/2020 to 18/11/2020 and compared to thirty-one healthy people (19 female and 12 male). Each participant was tested for H. pylori CagAAbs and TNF-α using an enzyme-linked immunosorbent assay (ELISA). The results have shown that there was a non-significant difference (p≥0.05) in the concentration of CagAantibodies in sera of patients with eye diseases except in the case of CSR (central serous chorioretinopathy), which was a significant difference (P≤0.05) compared to the control group. Also, the result
... Show MoreIn this paper the experimentally obtained conditions for the fusion splicing with photonic crystal fibers (PCF) having large mode areas were reported. The physical mechanism of the splice loss and the microhole collapse property of photonic crystal fiber (PCF) were studied. By controlling the arc-power and the arc-time of a conventional electric arc fusion splicer (FSM-60S), the minimum loss of splicing for fusion two conventional single mode fibers (SMF-28) was (0.00dB), which has similar mode field diameter. For splicing PCF (LMA-10) with a conventional single mode fiber (SMF-28), the loss was increased due to the mode field mismatch.
According to the theory of regular geometric functions, the relevance of geometry to analysis is a critical feature. One of the significant tools to study operators is to utilize the convolution product. The dynamic techniques of convolution have attracted numerous complex analyses in current research. In this effort, an attempt is made by utilizing the said techniques to study a new linear complex operator connecting an incomplete beta function and a Hurwitz–Lerch zeta function of certain meromorphic functions. Furthermore, we employ a method based on the first-order differential subordination to derive new and better differential complex inequalities, namely differential subordinations.
Mechanical degradation hampers the practical usage of polymers for turbulent drag reduction
application. Mechanical degradation refers to the chemical process in which the activation energy of
polymer chain scission is exceeded by mechanical action on the polymer chain, and bond rupture
occurs. When a water-soluble polymer and surfactant are mixed in water solution, the specific structures
(aggregates) are formed, in which polymer film is formed around micelle. In this work, Xanthan gum (XG) –
Sodium lauryl ether sulfate (SELS) complex formation and its effect on percentage viscosity reduction
(%VR) was studied. It was found that SELS surfactant reduced the mechanical degradation of XG much
more efficiently than th
The avoidance strategy of prey to predation and the predation strategy for predators are important topics in evolutionary biology. Both prey and predators adjust their behaviors in order to obtain the maximal benefits and to raise their biomass for each. Therefore, this paper is aimed at studying the impact of prey’s fear and group defense against predation on the dynamics of the food-web model. Consequently, in this paper, a mathematical model that describes a tritrophic Leslie-Gower food-web system is formulated. Sokol-Howell type of function response is adapted to describe the predation process due to the prey’s group defensive capability. The effects of fear due to the predation process are considered in the first two levels
... Show MoreThis work describes an experimental setup to evaluate the photodynamictoxicity of 650 nm diode laser and 532 nm Frequency-doubled Q-Switched Nd:YAG laser on the growth of Candida albicans as well as the potential fungicidal effect when combining the laser irradiation with specific photosensitizers namely methylene blue, toluidine blue, acridine orange and safranin O. In this study the findings showed that the number of colony-forming units per millilitre (CFU/ml) of C. albicans decreased with increasing exposure time. In particular in the case of the frequency doubled Nd:YAG laser combined with safranin O, the best lethal effect occurred at 11 minutes exposure time with 2.26 J/cm² energy density (89.18% reduction) in comparison with the
... Show MoreAbstract The aim of this study is preparing an intellectual map according to the feedback (verbally and writhingly ) in order to learn some skills of floor exercises in the women's artistic gymnastics , In addition to that the aim of this study defines on the impacts of intellectual map according to the feedback approach, and to identify the best group between the three groups study in the learning of skills approach in this study, the researchers used the experimental method, the subject of the study included on students second class in physical education and sport sciences , Baghdad University (2014-2015) , and divided into three groups for teaching skills which was under studied .The species used the specific manner by lot for selection
... Show More