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.
In the present article, mixed ligand metal (II) complexes have been synthesized with Schiff base (1E, 5Z, 6E)-1,7 bis (4-hydroxy-3- methoxyphenyl)-5-(3-hydroxyphenyl) imino) hepta-1,6-dien-3-one derived from Curcumin and 3-aminophenol as primary ligand and L-dopa as a secondary ligand. The Schiff base act as bidentate and arrange to the metals through the azomethine (C=N) nitrogen and (C=O) oxygen atom. The mode of bonding of the Schiff base has been affirmed on the infrared by the UV-Visible, 1H, and 13C NMR spectroscopic techniques. The magnetic susceptibility and the UV-Vis data of the complexes propose octahedral geometry around the central metal ion. The information appears that the complexes have the structure of [L-M-(L-dopa)] system
... Show MoreEncryption of data is translating data to another shape or symbol which enables people only with an access to the secret key or a password that can read it. The data which are encrypted are generally referred to as cipher text, while data which are unencrypted are known plain text. Entropy can be used as a measure which gives the number of bits that are needed for coding the data of an image. As the values of pixel within an image are dispensed through further gray-levels, the entropy increases. The aim of this research is to compare between CAST-128 with proposed adaptive key and RSA encryption methods for video frames to determine the more accurate method with highest entropy. The first method is achieved by applying the "CAST-128" and
... Show MoreThe second most commonly diagnosed cancer is colorectal cancer (CRC) is in female. The levels of progranulin, obestatin and liver enzymes including ALT, AST and ALP were measured in forty five sera in female patients suffering from CRC before chemotherapy initiation treatment as G1, G2 after first chemotherapy cycle and G3 after second chemotherapy cycle compared with thirty female as a healthy control G4. Results showed a high significant increased in progranulin concentration and a high significant decrease in obestatin in G2 than other groups. The correlation between progranulin and ALP was a significant negative (-ve) relation while obestatin with AST gave a significant positive (+ve) correlation in G. The results also showed non signif
... Show MoreBack ground: Several devices with different physical bases have been developed for the clinical measurement of corneal thickness, they classified into 4 categories: Scheimpflug photography based, Slit –Scanning topography, optical coherence tomography (OCT) based and ultrasound (US) based.Objective:To evaluatethe precision of the new Scheimpflug –Placido disc corneal topography in measurement of corneal thickness and to compare the measured values with that obtained by US pachymetry.Methods: Setting of this study is Lasik center in Eye Specialty Private Hospital. Baghdad. Iraq.Eyes of healthy subjects were examined with the Sirius topography.3 consecutive measurements of central (CCT)and thinnest (TCT) corneal thicknesses were obtain
... Show MoreIt is shown that pure and 3% boron doped a-Si0.1Ge0.9:H and a-Si0.1Ge0.9:N thin films
could be prepared by flash evaporation processes. The hydrogenation and nitrogenation
are very successful in situ after depositing the films. The FT-IR analysis gave all the
known absorbing bonds of hydrogen and nitrogen with Si and Ge.
Our data showed a considerable effect of annealing temperature on the structural and
optical properties of the prepared films. The optical energy gap (Eopt.) of a-Si0.1Ge0.9
samples showed to have significant increase with annealing temperature (Ta) also the
refractive index and the real part of dielectric constant increases with Ta, however the
extinction coefficient and imaginary part of dielect
Abstract
This research aims to define the roles of auditors by clarifying the concept and risks of cyber security in protecting information and financial data in economic units. Najaf, Babylon and Karbala, then the results were analyzed and the results were presented and analyzed to show that adopting cyber security improves the quality of reports Finance through what it achieves in displaying information with credibility and transparency, in a way that suits the needs of users, and cyber security has a role in managing economic resources more effectively to obtain benefits that would have been lost in the event of an
... Show MoreAbstract\
preparation process of performance evaluation in organizations is of extreme importance, and under development in organizations and the opening of markets and technological developments in the industry and heightened competition among industrial organizations imposed systems are built for performance give a clear picture about performance and competition, And centered research problem in answering the following questions: Is performance evaluation system is available in Wasit State Company for Textile Industries(Research sample ), This research aims to assess the performance of policies and programs in the company, according to guide performance evaluation of programs and policies prepared by the Dutch Cou
... Show More