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 study had succeeded in producing a new graphical representation of James abacus called nested chain abacus. Nested chain abacus provides a unique mathematical expression to encode each tile (image) using a partition theory where each form or shape of tile will be associated with exactly one partition.Furthermore, an algorithm of nested chain abacus movement will be constructed, which can be applied in tiling theory.
The aim of this study is to identify the effect of enabling the effectiveness of the work of the audit committees in private commercial banks and to identify the extent of awareness of the importance of empowerment in the work of these committees, especially as it is known that these committees, especially the inspection committees that go to private banks and from various sources including committees of the Central Bank of Iraq Committees of the Securities Commission and finally committees of the external audit offices, through an analysis of the determinants of empowerment in the performance of the most important work of the audit committees, namely: supervising the process of preparing reports, supervising the system of intern
... Show MoreThis work focuses on the preparation of pure nanocrystalline SnO2 and SnO2:Cu thin films on cleaned glass substrates utilizing a sol-gel spin coating and chemical bath deposition (CBD) procedures. The primary aim of this study is to investigate the possible use of these thin films in the context of gas sensor applications. The films underwent annealing in an air environment at a temperature of 500 ◦C for duration of 60 minutes. The thickness of the film that was deposited may be estimated to be around 300 nm. The investigation included an examination of the structural, optical, electrical, and sensing characteristics, which were explored across various preparation circumstances, specifically focusing on varied
... Show MoreWith the revolutionized expansion of the Internet, worldwide information increases the application of communication technology, and the rapid growth of significant data volume boosts the requirement to accomplish secure, robust, and confident techniques using various effective algorithms. Lots of algorithms and techniques are available for data security. This paper presents a cryptosystem that combines several Substitution Cipher Algorithms along with the Circular queue data structure. The two different substitution techniques are; Homophonic Substitution Cipher and Polyalphabetic Substitution Cipher in which they merged in a single circular queue with four different keys for each of them, which produces eight different outputs for
... Show MoreThis work bases on encouraging a generous and conceivable estimation for modified an algorithm for vehicle travel times on a highway from the eliminated traffic information using set aside camera image groupings. The strategy for the assessment of vehicle travel times relies upon the distinctive verification of traffic state. The particular vehicle velocities are gotten from acknowledged vehicle positions in two persistent images by working out the distance covered all through elapsed past time doing mollification between the removed traffic flow data and cultivating a plan to unequivocally predict vehicle travel times. Erbil road data base is used to recognize road locales around road segments which are projected into the commended camera
... Show Moren this study, data or X-ray images Fixable Image Transport System (FITS) of objects were analyzed, where energy was collected from the body by several sensors; each sensor receives energy within a specific range, and when energy was collected from all sensors, the image was formed carrying information about that body. The images can be transferred and stored easily. The images were analyzed using the DS9 program to obtain a spectrum for each object,an energy corresponding to the photons collected per second. This study analyzed images for two types of objects (globular and open clusters). The results showed that the five open star clusters contain roughly t
... Show More