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 developments accelerated in technology and rapid changes in the environment and increase numbers industrial countries and different desires and requirements of customers, lead to be produced in large quantities is not feasible due to changes listed above as well as the need to product variety and change in tastes and desires of consumers, all above led not to enable companies to discharge their products in the case of mass production and created the need to devise ways and new methods fit with the current situation, and accounting point no longer the traditional accounting systems able to meet the requirements needed by the companies to make decisions and know where waste and loss of resources resulting to invent new style away from
... Show MoreDust storms are typical in arid and semi-arid regions such as the Middle East; the frequency and severity of dust storms have grown dramatically in Iraq in recent years. This paper identifies the dust storm sources in Iraq using remotely sensed data from Meteosat-spinning enhanced visible and infrared imager (SEVIRI) bands. Extracted combined satellite images and simulated frontal dust storm trajectories, using the Hybrid Single Particle Lagrangian Integrated Trajectory (HYSPLIT) model, are used to identify the most influential sources in the Middle East and Iraq. Out of 132 dust storms in Iraq during 2020–2023, the most frequent occurred in the spring and summer. A dust source frequency percentage map (DSFPM) is generated using ArcGIS so
... Show MoreThe photo-electrochemical etching (PECE) method has been utilized to create pSi samples on n-type silicon wafers (Si). Using the etching time 12 and 22 min while maintaining the other parameters 10 mA/cm2 current density and HF acid at 75% concentration.. The capacitance and resistance variation were studied as the temperature increased and decreased for prepared samples at frequencies 10 and 20 kHz. Using scanning electron microscopy (SEM), the bore width, depth, and porosity % were validated. The formation of porous silicon was confirmed by x-ray diffraction (XRD) patterns, the crystal size was decreased, and photoluminescence (PL) spectra revealed that the emission peaks were centered at 2q of 28.5619° and 28.7644° for et
... Show MoreIn this research, cyclic compounds derived from 2- furfural mercaptan (oxazole, triazoles) were synthesized, and their biological efficacy was measured and compared with standard drugs. Also, their effectiveness as anti-oxidant was measured and compared with ascorbic acid as a standard substance. Some of the synthesized compounds were deduced with good efficacy. © 2021 Sami Publishing Company. All rights reserved
In this paper, the class of semi
Abstract
This research aims to study and improve the passivating specifications of rubber resistant to vibration. In this paper, seven different rubber recipes were prepared based on mixtures of natural rubber(NR) as an essential part in addition to the synthetic rubber (IIR, BRcis, SBR, CR)with different rates. Mechanical tests such as tensile strength, hardness, friction, resistance to compression, fatigue and creep testing in addition to the rheological test were performed. Furthermore, scanning electron microscopy (SEM)test was used to examine the structure morphology of rubber. After studying and analyzing the results, we found that, recipe containing (BRcis) of 40% from th
... Show MoreThis studies p- CuO / n - Si hete-rojunction was deposited by high vacuum thermal evaporation of Copper subjected to thermal oxidation at 300 oC on silicon. Surface morphology properties of The optical properties concerning the transmission spectra were studies for prepared thin films. this structure have been studied. XRD anaylsis discover that the peak at (𝟏𝟏𝟏-) and (111) plane are take over for the crystal quality of the CuO films. The band gap of CuO films is found to be 1.54 eV. The average grain size of is measured from AFM analysis is around 14.70 nm. The responsivity photodetector after deposited CuO appear increasing in response
Ag2O (Silver Oxide) is an important p-type (in chasm to most oxides which were n-type), with a high conductivity semiconductor. From the optical absorbance data, the energy gap value of the Ag2O thin films was 1.93 eV, where this value substantially depends on the production method, vacuum evaporation of silver, and optical properties of Ag2O thin films are also affected by the precipitation conditions. The n-type and p-type silicon substrates were used with porous silicon wafers to precipitate ±125 nm, as thick Ag2O thin film by thermal evaporation techniques in vacuum and via rapid thermal oxidation of 400oC and oxidation time 95 s, then characterized by measurement of
... Show Morethe research goal is preparing a list of standard criteria and quality controls for information technology applications to serve the Holy Quran.
To achieve this goal, the researcher has built a list of criteria according to the following steps:
First - identify the key areas covered by the whole list which are:
1 – Standards of system building and implementing with the operating screens.
2 – Standards of display forms including audio and video presentation.
3 – Standards which are related to the program philosophy.
4 - Standards which are related to the program objectives.
... Show More