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.
By reading the book (Endless Forms Most Beautiful: The New Science of Evo Devo) by Sean B. Carroll, new horizons opened up about the nature of the formation of the living organism. Although he presented the idea that the artist was influenced by the material assets of nature in his holographic art formations, the new science of Evo-Devo (Evolutionary Developmental Science) provided models worth standing on when comparing the similarity of the formation of living organisms on the one hand, and the formation of works of art with holographic organic bodies on the other. But the excitement lies in the fact that the formation of living natural organisms is often driven by subtle intelligent mechanisms that are different from the mechanisms us
... Show MoreThe research topic was chosen as a result of the importance of human resource in business organizations in general and the industrial process in particular. Without the human resource, business organizations cannot continue and achieve success and excellence, and the research problem has been diagnosed in the lack of sales of General Cement Company’s northern products, despite their distinctiveness, standing, and reputation in The market and its products with standard specifications, and through this problem, the following questions were raised: &nbs
... Show MoreIn the knowledge society, artificial intelligence (AI) forms a cornerstone of global education. This quasi-experimental study examines the impact of an Intelligent Adaptive Learning Strategy (IALS) on flexible thinking (FT) and academic achievement among 60 3rd-year undergraduate students at the College of Education/University of Baghdad (experimental n = 30; control n = 30). The IALS was implemented via an AI-supported educational platform, while the control group received conventional instruction. Post-test intervention assessments included an FT test (10 items, content validity = 0.89, Cronbach’s α = 0.87) and an achievement test (10 objective items, α = 0.85). Results revealed statistically significant superiority of the exp
... Show Morestudy the effect of radiation microwave (MW) in inhibition the growth of some types of bacteria in a minced meat and barker were exposed to MW for different times included (0, 10, 20, 30 and 40) sec.The results showed a high inhibition rate for 40 sec, reached to 100%. It is the other side studied the effect of microwave radiation against four types of bacteria included (Staphylococcus aureus, Escherichia coli, Proteus mirabilis and Klebsiella spp), when were exposed to for (0, 5, 10, 20, 30 and 40) sec the inhibition ratio reached to 100% in each of the Proteus mirabilis and Klebsiella spp at 30 sec and Staphylococcus aureus and Escherichia coli at 40sec. using MW in the sterilization media, such as Nutrient agar, Macconkey agar and Man
... Show MoreAir pollution refers to the release of pollutants into the air that are detrimental to human health and the planet as a whole.In this research, the air pollutants concentration measurements such as Total Suspended Particles(TSP), Carbon Monoxides(CO),Carbon Dioxide (CO2) and meteorological parameters including temperature (T), relative humidity (RH) and wind speed & direction were conducted in Baghdad city by several stations measuring numbered (22) stations located in different regions, and were classified into (industrial, commercial and residential) stations. Using Arc-GIS program ( spatial Analyses), different maps have been prepared for the distribution of different pollutant
Survival analysis is widely applied in data describing for the life time of item until the occurrence of an event of interest such as death or another event of understudy . The purpose of this paper is to use the dynamic approach in the deep learning neural network method, where in this method a dynamic neural network that suits the nature of discrete survival data and time varying effect. This neural network is based on the Levenberg-Marquardt (L-M) algorithm in training, and the method is called Proposed Dynamic Artificial Neural Network (PDANN). Then a comparison was made with another method that depends entirely on the Bayes methodology is called Maximum A Posterior (MAP) method. This method was carried out using numerical algorithms re
... Show MoreTonsillitis is an infection or inflammation of the tonsils. Tonsillitis classified as acute, chronic (recurrent) and complication. Most often it is caused by a virus, but it may also be caused by the same bacteria that cause strep throat. According to numbers that data shows, the virus is more common, about 77.49% than bacterial which about 71.45% then fungal about 9.6%. The principal symptom of tonsillitis is a sore throat fever, swollen lymph nodes, nasal congestion, difficulty in swallowing and headache may also occur, according to if symptoms of tonsillitis are known or not, data show that they know about (119 of people) (79.3) this result is greater than unknown which was about 31 (20.7).