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.
This paper aims to investigate the flexural behavior of reinforced concrete beams considering fire resistance by adding Lightweight Expanded Clay Aggregates (LECA) to the concrete mix as partial coarse aggregate replacement. LECA is a type of porous clay with a uniform pore structure with fine, closed cells and hard, tightly sintered skin. The experimental work comprised four reinforced self-compacted concrete beams. All the specimens were identical in their geometrical layout of 1600×240×200 mm, reinforcement details, and support condition (simply supported). For all the beams, the main reinforcement was provided by two bars, each having a diameter of 12 mm, while a bar of 6 mm diameter was employed for the top and shear reinforc
... Show MoreBackground: A worldwide health epidemic, type 2 diabetes mellitus was significantly influenced by chronic inflammation, which led to increased insulin resistance (IR). The most widely practiced form of therapy used to control musculoskeletal pain in people with diabetes is non-steroidal anti-inflammatory drugs (NSAIDs), which provide their action by inhibiting cyclooxygenase enzyme (COX). COX1, COX2, and COX3 are distinct isoforms of the cyclooxygenase enzyme. The potential anti-inflammatory benefits of cyclooxygenase-2 (COX-2) inhibitors, both selective and non-selective, have been investigated concerning the management of type 2 diabetes patients. Objective: the purpose of this research is to explore the impact of highly selective
... Show MoreABSTRACT Background: resin cement type and intraoral temperature fluctuations may affect the fracture performance of successful zirconia restorations. To fill this gap, the purpose of this study is to evaluate and compare the influence of thermocycling on fracture resistance and mode of failure of monolithic zirconia crowns luted with Rely X™ U200 and BreezeTMself-adhesive resin cements as well as imply the effect of adding 2 % of polylysine (PLS) to these cements. Materials: 64 maxillary premolars were milled out of zirconia blocks using CAD/CAM milling system. They were divided into four groups (n = 16) according to the cement type. Four different resin cements were used (RelyXTMU200, Breeze™, RelyX™ U200 with 2 % PLS
... Show MoreObjectives: to evaluate patient knowledge with hemodialysis and to determine the effectiveness of Self-regulation Fluid Program on Patients with hemodialysis self-efficacy for fluid adherence in Al-Diwaniyah Teaching Hospital.
Methodology: A quasi-experimental design (two group design: pre-test and post-test) was used. This study was conducted in Al-Diwaniya Teaching Hospital for the period from (15th of October 2018 to 20th of May 2019) on a non-probability (purposive) sample consisting of (60 patients) treatment in hemodialysis units. A questionnaire was built as a data collection tool and consisted of four parts:
First part: Demographic characteristics of the pati
... Show More
Abstract
This study aims to identify the degree to which the first cycle teachers use different feedback patterns in the e-learning system in addition to the differences in the degree of use according to specialization, teaching experience, and in-service training in the field of classroom assessment, as well as the interaction between them. The study sample consisted of (350) female teachers of the first cycle in government schools in Muscat Governorate for the academic year 2020/2021. The study used a questionnaire that contained four different patterns of feedback, which are reinforcement, informative, corrective, and interpretive feedback. The psychometric properties of the que
... Show MoreIn this research PbS and PbS:Cu films were prepered with thicknesses (0.85±0.05)?m and (0.55±0.5)?m deposit on glass and silicon substrate respectively using chemical spray pyrolysis technique with a substrate temperature 573K, from lead nitrate salt, thiourea and copper chloride. Using XRD we study the structure properties for the undoped and doped films with copper .The analysis reveals that the structure of films were cubic polycrystalline FCC with a preferred orientation along (200) plane for the undoped films and 1% doping with copper but the orientation of (111) plane is preferred with 5% doping with the rest new peaks of films and appeared because of doping. Surface topography using optical microscope were be checked, it was found
... Show MoreBackground: Cerebral palsy is non-progressive disorder of posture or movement due to a lesion of the developing brain. It is the commonest physical disability in childhood. Objective: To study the clinical, neurological abnormalities, prevalence of convulsion (epilepsy) & to asses the value of CT scans of brain in patients with cerebral palsy.
JM Karhoot, AA Noaimi, WF Ahmad, The Iraqi Postgraduate Medical Journal, 2012 - Cited by 7
Using the Internet, nothing is secure and as we are in need of means of protecting our data, the use of passwords has become important in the electronic world. To ensure that there is no hacking and to protect the database that contains important information such as the ID card and banking information, the proposed system stores the username after hashing it using the 256 hash algorithm and strong passwords are saved to repel attackers using one of two methods: -The first method is to add a random salt to the password using the CSPRNG algorithm, then hash it using hash 256 and store it on the website. -The second method is to use the PBKDF2 algorithm, which salts the passwords and extends them (deriving the password) before being ha
... Show More