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.
The data preprocessing step is an important step in web usage mining because of the nature of log data, which are heterogeneous, unstructured, and noisy. Given the scalability and efficiency of algorithms in pattern discovery, a preprocessing step must be applied. In this study, the sequential methodologies utilized in the preprocessing of data from web server logs, with an emphasis on sub-phases, such as session identification, user identification, and data cleansing, are comprehensively evaluated and meticulously examined.
Green synthesis of bimetallic nanoparticles of Fe/Ni (G-Fe/Ni-NPs) and zeolite-5A supported (G-Z-Fe/Ni-NPs) as heterogeneous Fenton-like oxidation for the decolourisation of reactive red 120-dye (RR120) from the aqueous medium using green tea extract as a reducing agent. Zeolite-5A from local kaolin is prepared and characterised using the hydrothermal method and is used as a supporting material for Fe/Ni-NPs. (SEM), (EDX), (AFM), (XRD), (FT-IR), (BET). Its zeta potential were used to characterise G-Fe/Ni-NPs and G-Z-Fe/Ni-NPs. The decolourisation efficiency (Ed) of the RR120-dye using a heterogeneous Fenton-like for G-Fe/Ni-NPs and G-Z-Fe/Ni-NPs is 99.8% and 99.9%, respectively, under the optimum conditions: [H2O2] = 20 and 1 mmol/L
... Show MoreFacing the Iraqi economy, a number of economic challenges that threaten the future of Iraq and the security of economic, political and social, such as poverty, unemployment, inflation and the dilapidated infrastructure and rising production costs and administrative and financial corruption, environmental pollution, water problems and the deterioration of agricultural and industrial production, etc., and over the seriousness of these challenges, they are intertwined and overlapping and growing worse, without the corresponding adoption of state strategies that will develop appropriate solutions and appropriate to resolve those challenges because of concern the subject of security and terrorism, which requires the development of an
... Show MoreThe current research aims at: - Identifying the role played by the leadership in empowerment and organizational learning abilities and their reflection on the knowledge capital, and the extent to which these concepts can be applied effectively at Wasit University. The problem of research .... In a series of questions: The most important is that the dimensions leadership empowerment and distance learning organizational capacity correlation relationship and impact and significant statistical significance with the capital knowledge.
To understand the nature of the relationship and the impact between the variables, leadership was adopted by empowerment as the fir
... Show MoreWireless Body Area Sensor Networks (WBASNs) have garnered significant attention due to the implementation of self-automaton and modern technologies. Within the healthcare WBASN, certain sensed data hold greater significance than others in light of their critical aspect. Such vital data must be given within a specified time frame. Data loss and delay could not be tolerated in such types of systems. Intelligent algorithms are distinguished by their superior ability to interact with various data systems. Machine learning methods can analyze the gathered data and uncover previously unknown patterns and information. These approaches can also diagnose and notify critical conditions in patients under monitoring. This study implements two s
... Show MoreAt the local level in the early 1990s first established an Islamic bank 1993 is called Iraqi Islamic bank and was established by Dr. abdullatef hmim and with capital of 200 Million Iraqi dinars .
After 2003 began the attention of Iraqis tend to create Islamic banks until now up to 30 banks in the present , He confirmed Mr. Governor of the Iraqi Central Bank recently that there is more than 40 iintroduction licence to establish an Islamic bank in Iraq , That banking activity which is compatible with Sharia law widely welcomed by savers and those in need for more development and modernization in order to gain a greater share within the Iraqi market , and allowed the Iraqi Central Bank in recent years
... Show MoreA total of 533 specimens were collected in survey of Brachyceran species from different regions of Iraq during February to November 2014 .This study was reported 16 species belonging to 13 genera and 7 families, the results showed that Dicranosepsis Duda, 1926 (Family; Sepsidae) is recorded the genus for the first time in Iraq.
The vast advantages of 3D modelling industry have urged competitors to improve capturing techniques and processing pipelines towards minimizing labour requirements, saving time and reducing project risk. When it comes to digital 3D documentary and conserving projects, laser scanning and photogrammetry are compared to choose between the two. Since both techniques have pros and cons, this paper approaches the potential issues of individual techniques in terms of time, budget, accuracy, density, methodology and ease to use. Terrestrial laser scanner and close-range photogrammetry are tested to document a unique invaluable artefact (Lady of Hatra) located in Iraq for future data fusion sc
This investigation presents an experimental and analytical study on the behavior of reinforced concrete deep beams before and after repair. The original beams were first loaded under two points load up to failure, then, repaired by epoxy resin and tested again. Three of the test beams contains shear reinforcement and the other two beams have no shear reinforcement. The main variable in these beams was the percentage of longitudinal steel reinforcement (0, 0.707, 1.061, and 1.414%). The main objective of this research is to investigate the possibility of restoring the full load carrying capacity of the reinforced concrete deep beam with and without shear reinforcement by using epoxy resin as the material of repair. All be
... Show More