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.
Introduction: Cardiovascular diseases are the main cause of death among type 2 diabetic patients. Higher levels of plasminogen activator urokinase receptor have been found to predict morbidity and mortality across acute and chronic diseases in the common populace. This study aims to explore the role of serum plasminogen activator urokinase receptor levels as a cardiometabolic risk factor among type 2 diabetic Iraqi patients. Methods: Seventy type 2 diabetic patients (40 male and 30 female) (mean age: 46.20±7.56 years) participated in this study; 35 patients were with cardiovascular disease and 35 were without cardiovascular disease; their ages range was 40-55 years. In addition, 30 individuals who apparently healthy were selected a
... Show MoreThis research deals with the study of top soil electrical conductive regions located within Baghdad City. The research included measuring the dissolved soil material extraction Electrical Conductivity (EC) with an aqueous solution for the top (0-30 cm) soil layer of the study area. As the electrical conductivity values increase by increasing the amount of dissolved salts in principle, we can consider that the aim of this research is to predict the amount and distribution of (soil contamination with salts) which is represented by the (Salt Index), this factor calculated for each soil representative sample taken from the region with a depth of (30 cm). Laboratory (EC) test values measured by the use of solutions (EC) digital meter for the ex
... Show MoreReliability analysis methods are used to evaluate the safety of reinforced concrete structures by evaluating the limit state function 𝑔(𝑋𝑖). For implicit limit state function and nonlinear analysis , an advanced reliability analysis methods are needed. Monte Carlo simulation (MCS) can be used in this case however, as the number of input variables increases, the time required for MCS also increases, making it a time consuming method especially for complex problems with implicit performance functions. In such cases, MCS-based FORM (First Order Reliability Method) and Artificial Neural Network-based FORM (ANN FORM) have been proposed as alternatives. However, it is important to note that both MCS-FORM and ANN-FORM can also be time-con
... Show MoreBackground: The type of dental implant surface is one of many factors that determine the success of implant restoration. This study aimed to study the effect of mixture of nano titanium oxide with nanohydroxyapatite coating of screw shaped CPTi dental implant on bond strength at bone implant interface by torque removal test related to two healing periods (2 and 6 weeks). Materials and methods: Dip coating process was performed to get an even coating layer on CPTi screws. X-ray diffraction (XRD) analysis and microscopical examination were performed on the coating surfaces of the CPTi. The tibia of 10 white New Zealand rabbits was chosen as implantation sites. The tibia of each rabbit received two screws, one was coated with mixture of nanoT
... Show MoreThe aim of this study was to assess the effectiveness of listening to music or Quran in reducing cancer patients’ anxiety before chemotherapy administration. Reducing anxiety in people with cancer, prior to chemotherapy administration, is a crucial goal in nursing care.
An experimental comparative study was conducted.
A simple randomization sampling method was applied. Two hundred thirty‐eight people with cancer who underwent chemotherapy were participated. They are assigned as Quran, music and control groups.
The current study aimed to identify the difficulties faced by the student in mathematics and possible proposals to address these difficulties. The study used a descriptive method also used the questionnaire to collect data and information were applied to a sample of (163) male and female teachers. The results of the study found that the degree of difficulties in learning mathematics for the fifth and sixth grades is high for some paragraphs and intermediate for other paragraphs, included the student's field. The results also revealed that there were no statistically significant differences at the level of significance (α = 0.05) between the responses of the members of the study sample from male and female teachers to the degree of diffi
... Show MoreAbstract: The research covered five chapters: So, the first chapter definition of the research is from the introduction to the research and its importance, as the importance of the research lies in an expression of the reality of e-learning as it is one of the new patterns of the educational process and its role in enhancing communication and interconnectedness between the learners from the students ’point of view Physical Education and Sports Sciences for Girls, University of Baghdad, as for the problem The research was, and through the researcher’s acquaintance with many previous studies, references and sources, and being a student at the College of Physical Education and Sports Sciences - University of
... Show More