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.
Given the paucity and toxicity of available drugs for leishmaniasis, coupled with the advent of drug resistance, the discovery of new therapies for this neglected tropical disease is recognised as being of the utmost urgency. As such antimicrobial peptides (AMPs) have been proposed as promising compounds against the causative Leishmania species, insect vector-borne protozoan parasites. Here the AMP temporins A, B and 1Sa have been synthesised and screened for activity against Leishmania mexicana insect stage promastigotes and mammalian stage amastigotes, a significant cause of human cutaneous disease. In contrast to previous studies with other species the activity of these AMPs against L. mexicana amastigotes was low. This suggests that ama
... Show MoreThis paper studies the behavior of axially loaded RC columns which are confined with carbon fiber reinforced polymers’ sheet (CFRP) and steel jackets (SJ). The study is based on twelve axially loaded RC columns tested up to failure. It is divided into three schemes based on its strengthening type; each scheme has four columns. The main parameters in this study were the compressive strength of the concrete and steel reinforcement ratio. Furthermore, the results of the experimental test showed a substantial enhancement in the column's load-carrying capacity. When compared to the original columns, the CFRP sheet had a significant effect on improving the ductility of the column by increasing the axial deformation by about 59.2 to 95.7
... Show MoreIn this paper, we investigate the connection between the hierarchical models and the power prior distribution in quantile regression (QReg). Under specific quantile, we develop an expression for the power parameter ( ) to calibrate the power prior distribution for quantile regression to a corresponding hierarchical model. In addition, we estimate the relation between the and the quantile level via hierarchical model. Our proposed methodology is illustrated with real data example.
This work consists of a numerical simulation to predict the velocity and temperature distributions, and an experimental work to visualize the air flow in a room model. The numerical work is based on non-isothermal, incompressible, three dimensional, k turbulence model, and solved using a computational fluid dynamic (CFD) approach, involving finite volume technique to solve continuity, momentum and energy equations, that governs the room’s turbulent flow domain. The experimental study was performed using (1/5) scaled room model of the actual dimensions of the room to simulate room air flow and visualize the flow pattern using smoke generated from burnt herbs and collected in a smoke generator to delivered through
... Show MoreABSTRACT: BACKGROUND: Estrogens has traditionally been known as the female hormone, but this idea has been challenged in early 1990’s and an essential physiological role for estrogen in male fertility was identified. Phytoestrogens are naturally occurring non-steroidal plant chemicals that can act like the female hormone estrogen. The herbs ( anise alfalfa and vervain ) chosen in this study contain phytoestrogens. OBJECTIVE: Previous studies demonstrated controversy of the effects of phytoestrogens on the rat testes .Hence, the present investigation was undertaken to investigate the influence of typical dose of herbs containing phytoestrogen on the rat testis. MATERIALS AND METHODS: Twenty-four apparently normal mature male rats we
... Show MoreThe adsorption of Malonic acid, Succinic acid, Adipic acid, and Azelaic acid from their aqueous solutions on zinc oxide surface were investigated. The adsorption efficiency was investigated using various factors such as adsorbent amount, contact time, initial concentration, and temperature. Optimum conditions for acids removal from its aqueous solutions were found to be adsorbent dose (0.2 g), equilibrium contact time (40 minutes), initial acids concentration (0.005 M). Variation of temperature as a function of adsorption efficiency showed that increasing the temperature would result in decreasing the adsorption ability. Kinetic modeling by applying the pseudo-second order model can provide a better fit of the data with a greater correla
... Show MoreIn the recent years the research on the activated carbon preparation from agro-waste and byproducts have been increased due to their potency for agro-waste elimination. This paper presents a literature review on the synthesis of activated carbon from agro-waste using microwave irradiation method for heating. The applicable approach is highlighted, as well as the effects of activation conditions including carbonization temperature, retention period, and impregnation ratio. The review reveals that the agricultural wastes heated using a chemical process and microwave energy can produce activated carbon with a surface area that is significantly higher than that using the conventional heating method.
Fuzzy logic is used to solve the load flow and contingency analysis problems, so decreasing computing time and its the best selection instead of the traditional methods. The proposed method is very accurate with outstanding computation time, which made the fuzzy load flow (FLF) suitable for real time application for small- as well as large-scale power systems. In addition that, the FLF efficiently able to solve load flow problem of ill-conditioned power systems and contingency analysis. The FLF method using Gaussian membership function requires less number of iterations and less computing time than that required in the FLF method using triangular membership function. Using sparsity technique for the input Ybus sparse matrix data gi
... Show More