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.
Over the past few decades, the surveying fieldworks were usually carried out based on classical positioning methods for establishing horizontal and vertical geodetic networks. However, these conventional positioning techniques have many drawbacks such as time-consuming, too costly, and require massive effort. Thus, the Global Navigation Satellite System (GNSS) has been invented to fulfill the quickness, increase the accuracy, and overcome all the difficulties inherent in almost every surveying fieldwork. This research assesses the accuracy of local geodetic networks using different Global Navigation Satellite System (GNSS) techniques, such as Static, Precise Point Positioning, Post Processing Kinematic, Session method, a
... Show MoreCopper selenide (Cu2Se) thin films were prepared by thermal evaporation at RT with thickness 500 nm. The heat-treating for (400 &500) K for the absorber layer has been investigated. This research includes, studying the structural properties of X-ray diffraction (XRD) that show the Cu2Se thin film (Cubic) and has a polycrystalline orientation prevalent (220). Moreover, studying the effect of annealing on their surface morphology properties by using Atomic Force Microscopy AFM. Optical properties were considered using the transmittance and absorbance spectra had been recorded when wavelength range (400 - 1000) nm in order to study the absorption coefficient and energy gap. It was found that these films had allowed direct transitio
... Show MoreThe human kidney is one of the most important organs in the human body; it performs many functions
and has a great impact on the work of the rest of the organs. Among the most important possible treatments is
dialysis, which works as an external artificial kidney, and several studies have worked to enhance the
mechanism of dialysate flow and improve the permeability of its membrane. This study introduces a new
numerical model based on previous research discussing the variations in the concentrations of sodium,
potassium, and urea in the extracellular area in the blood during hemodialysis. We simulated the differential
equations related to mass transfer diffusion and we developed the model in MATLAB Simu
In this research, we have added nano anatase TiO2 as a partial replacement of Portland cement by a weight percentage of (0.25 to 1%) for the development of properties for protection against bacteria. The control mix was made by using "the cement to sand" proportion about (1: 2.75) with the "water to cement" proportion of (0.5) to study the structure, porosity, water absorption, density, mechanical properties, as well as anti-bacterial behavior. Inspections have been done such as scanning electron microscopy (SEM), and atomic force microscope (AFM) for mortar. Experimental results showed that after the addition of Nano powders in cement mortar, the structural properties improved significantly with the development of hydration o
... Show MoreMedia and communication's research are varied in accordance to research approaches' variety which seeks to reach convergent social, psycholo
... Show MoreThis article describes how to predict different types of multiple reflections in pre-track seismic data. The characteristics of multiple reflections can be expressed as a combination of the characteristics of primary reflections. Multiple velocities always come in lower magnitude than the primaries, this is the base for separating them during Normal Move Out correction. The muting procedure is applied in Time-Velocity analysis domain. Semblance plot is used to diagnose multiples availability and judgment for muting dimensions. This processing procedure is used to eliminate internal multiples from real 2D seismic data from southern Iraq in two stages. The first is conventional Normal Move Out correction and velocity auto picking and
... Show MoreWellbore instability is one of the most common issues encountered during drilling operations. This problem becomes enormous when drilling deep wells that are passing through many different formations. The purpose of this study is to evaluate wellbore failure criteria by constructing a one-dimensional mechanical earth model (1D-MEM) that will help to predict a safe mud-weight window for deep wells. An integrated log measurement has been used to compute MEM components for nine formations along the studied well. Repeated formation pressure and laboratory core testing are used to validate the calculated results. The prediction of mud weight along the nine studied formations shows that for Ahmadi, Nahr Umr, Shuaiba, and Zubair formations
... Show MoreThe present study investigated the use of pretreated fish bone (PTFB) as a new surface, natural waste and low-cost adsorbent for the adsorption of Methyl green (MG, as model toxic basic dye) from aqueous solutions. The functional groups and surface morphology of the untreated fish bone (FB) and pretreated fish bone were characterized using Fourier transform infrared (FTIR), scanning electron microscopy (SEM) and Energy dispersive X-ray spectroscopy (EDS),respectively. The effect of operating parameters including contact time, pH, adsorbent dose, temperature, and inorganic salt was evaluated. Langmuir, Freundlich and Temkin adsorption isotherm models were studied and the results showed that the adsorption of basic dye followed Freundlich iso
... Show More