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.
ER Abbas, AA Jasim, Journal of Physical Education, 2023 - Cited by 1
Background: Masseter muscle is one of the most obvious muscles of mastication and considered as one indicator of jaw muscle activity. It has a major influence on the transverse growth of the midface and the vertical growth of the mandible. This study undertaken to determine the role of cephalometric analysis for discrimination between Cl I and Cl III skeletal relationships, determine the role of ultrasonography in determination of masseter muscle thickness, compare masseter muscle thickness between Cl I and Cl III skeletal relationships, and determine the effect of gender on masseter muscle thickness. Material and Method: The sample of the current study consisted of 70 Iraqi subjects 40 males and 30 females with age ranging 18-25 years. The
... Show MoreВ статье рассматриваются понятие наглядности и ее классификация, эффективное применение принципа наглядности, возможность и преимущество использования наглядности и современные инструменты ее реализации в ираке.
Вследствие глобальных изменений во всем мире изменилась и роль иностранного языка в системе образования. Современный человек, шагающий в ногу со временем, не мыслит себя без зна
... Show MoreBackground: The risk of antibiotics resistance (AR) increases due to excessive of antibiotics either by health care provider or by the patients.
Objective: The assessment of the self-medication Practice of over the counter drugs and other prescription drugs and its associated risk factor.
Subjects and Methods: Study design: A descriptive study was conducted from “20th December 2019 to 08th January 2021”. A pre validated and structured questionnaire in English and Urdu language was created to avoid language barrier including personal detail, reasons and source and knowledge about over the counter drugs and Antibiotics. Sample of the study was randomly selected.
... Show More
In the last years, the self-balancing platform has become one of the most common candidates to use in many applications such as flight, biomedical fields, industry. This paper introduced the simulated model of a proposed self-balancing platform that described the self–balancing attitude in (X-axis, Y-axis, or both axis) under the influence of road disturbance. To simulate the self-balanced platform's performance during the tilt, an integration between Solidworks, Simscape, and Simulink toolboxes in MATLAB was used. The platform's dynamic model was drawn in SolidWorks and exported as a STEP file used in the Simscape Multibody environment. The system is controlled using the proportional-integral-deriva
... Show MoreStenography is the art of hiding the very presence of communication by embedding secret message into innocuous looking cover document, such as digital image, videos, sound files, and other computer files that contain perceptually irrelevant or redundant information as covers or carriers to hide secret messages.
In this paper, a new Least Significant Bit (LSB) nonsequential embedding technique in wave audio files is introduced. To support the immunity of proposed hiding system, and in order to recover some weak aspect inherent with the pure implementation of stego-systems, some auxiliary processes were suggested and investigated including the use of hidden text jumping process and stream ciphering algorithm. Besides, the suggested
... Show MoreImproving speaking skills of Iraqi EFL students was the main purpose of the current research. Thirty EFL students were selected as the research participants for achieving this aim. All students completed the pretest and then spent the next 25 weeks meeting for 90 minutes each to present their nine lectures, answer difficult questions, and get feedback on their use of language in context. Progressive-tests, posttests and delayed post-tests followed every three courses. The researcher utilized SPSS 22 to anal Analyze the data descriptively and inferentially after doing an ANOVA on repeated measurements. It has been shown that using the ideas of sociocultural theory in the classroom has an important and positive impact on students of
... Show MoreInduced EF is among the most important of advanced oxidation processes (AOPs) It was employed to treat different kinds of wastewater. In the present review, the types and mechanism of induced EF were outlined. Parameters affecting this process have been mentioned with details. These are current density, pH, H2O2 concentration, and time. The application of induced electro Fenton in various sectors of industries like textile, petroleum refineries, and pharmaceutical were outlined. The outcomes of this review demonstrate the vital role of induced EF in treatment of wastewater at high efficiency and low cost in contrast with conventional technique
