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.
Fiber-to-the-Home (FTTH) has long been recognized as a technology that provides future proof bandwidth [1], but has generally been too expensive to implement on a wide scale. However, reductions in the cost of electro-optic components and improvements in the handling of fiber optics now make FTTH a cost effective solution in many situations. The transition to FTTH in the access network is also a benefit for both consumers and service providers because it opens up the near limitless capacity of the core long-haul network to the local user. In this paper individual passive optical components, transceivers, and fibers has been put together to form a complete FTTH network. Then the implementation of the under construction Baghdad/Al
... Show MoreDBN Rashid, Rimak International Journal of Humanities and Social Sciences, 2020
Judicial jurisprudence is one of the important legal solutions to address the shortcomings of legislation. Throughout its long history, human societies have known many cases in which the judge finds himself facing a legislative vacuum in addition to civil legal texts that are difficult for the judge to implement due to ambiguity or contradiction, which requires diligence. To rule on resolving disputes before him in order not to deny justice, but the judge in his jurisprudence was not absolute, but rather bound by certain controls represented by observing the wisdom of legislation on the one hand and taking into account the nature of the texts on the other side, and from here this research came to shed light on the jurisprudence and its cont
... Show MorePV connected systems are worldwide installed because it allows consumer to reduce energy consumption from the electricity grid. This paper presents the results obtained from monitoring a 1.1 kWp. The system was monitored for nine months and all the electricity generated was fed to the fifth floor for physics and renewable energy building 220 V, 50 Hz. Monthly, and daily performance parameters of the PV system are evaluated which include: average generated of system Ah per day, average system efficiency, solar irradiation around these months. The average generated kWh per day was 8 kWh/day, the average solar irradiation per day was 5.6 kWh/m2/day, the average inverter efficiency was 95%, the average modules efficien
... Show MoreFe, Co and Sb nanopowders were fruitfully prepared by electrical wire explosion method in Double distilled and de-ionized water (DDDW) media. The formation of iron, cobalt and antimony (FeCoSb) alloy nanopowder was monitored by X-ray diffraction. The x-ray diffraction pattern indicates that there are iron, cobalt and antimony peaks. Optical properties of this alloy nanoparticles were characterized by UV-Visible absorption spectra. The absorption peak position is shifted to the lower wavelengths when the current increases. That means the mean size of the nanoparticles controlled by changing the magnitude of the current. The surface morphological analysis is carried out by employing Scanning Electron Microscope (SEM). Particles with varies
... Show MoreMost of the Internet of Things (IoT), cell phones, and Radio Frequency Identification (RFID) applications need high speed in the execution and processing of data. this is done by reducing, system energy consumption, latency, throughput, and processing time. Thus, it will affect against security of such devices and may be attacked by malicious programs. Lightweight cryptographic algorithms are one of the most ideal methods Securing these IoT applications. Cryptography obfuscates and removes the ability to capture all key information patterns ensures that all data transfers occur Safe, accurate, verified, legal and undeniable. Fortunately, various lightweight encryption algorithms could be used to increase defense against various at
... Show MoreThere are many ways to communicate a story, one can tell it, sing it, act it, write it, film it and can draw it. Each form of storytelling has its unique characteristics that separate it from the other forms. These distinguishing characteristics of each form does not indicate the superiority of one form over the other, on the contrary, it indicates the human ability to be creative on many levels. As students of literature, only the written form of a story is taken into account. One reason for the negligence of other forms of storytelling might be that singing and filming a story is not related to literature. But there has been a form of storytelling that uses words along with pictures which eventually became a genre of novels called grap
... Show More
