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.
Abstract
The extremes effects in parameters readings which are BOD (Biological Oxygen Demands) and DO(Dissolved Oxygen) can caused error estimating of the model’s parameters which used to determine the ratio of de oxygenation and re oxygenation of the dissolved oxygen(DO),then that will caused launch big amounts of the sewage pollution water to the rivers and it’s turn is effect in negative form on the ecosystem life and the different types of the water wealth.
As result of what mention before this research came to employees Streeter-Phleps model parameters estimation which are (Kd,Kr) the de oxygenation and re oxygenation ratios on respect
... Show MoreIn this paper, effective slab width for the composite beams is investigated with special emphasis on the effect of web openings. A three dimensional finite element analysis, by using finite element code ANSYS, is employed to investigate shear lag phenomenon and the resulting effective slab width adopted in the classical T-beam approach. According to case studies and comparison with limitations and rules stipulated by different standards and codes of practice it is found that web openings presence and panel proportion are the most critical factors affecting effective slab width, whereas concrete slab thickness and steel beam depth are less significant. The presence of web opening reduces effective slab width by about 21%.
... Show More|
The international organizations showed their interest in the marketable public relations, with their activities, means and strategies, they play a crucial role in marketing products, services and ideas of the institution. They are considered to be the link between the company and its public, They are responsible for presenting the institution to the public, with honest transmission of the information. This gives a good impression to the institution, in a way the institution and its products become consistent with the needs and interests of the public. Based on this, the research aims to identify the strategies used by the marketable public relations internationally. The r |
Background: During acrylic resin processing, the mold must be separated from the surface of the gypsum to prevent liquid resin from penetrating into the gypsum, and water from the gypsum seeping into the acrylic resin. For many years, tin foil was the most acceptable separating medium, and because it's difficult to apply, a tin-foil substitute is used. In this study, olive oil is used as an alternative to tin foil separating medium for first time, and evaluating its effect as a separating medium on some mechanical properties such as (indentation hardness and transverse strength) of acrylic resins denture base comparing it with those processed using tin-foil and tin foil substitute such as (cold mold seal) separating medium. Materials and M
... Show MoreIn this research work, a new type of concrete based on sulfur-melamine modification was introduced, and its various properties were studied. This new type of concrete was prepared based on the sulfur-melamine modification and various ingredients. The new sulfur-melamine modifier was fabricated, and its fabrication was confirmed by IR spectroscopy and TG analysis. The surface morphology resulted from this modifier was studied by SEM and EDS analysis. The components ratios in concrete, chemical and physical characteristics resulted from sulfur-melamine modifier, chemical and corrosion resistance of concrete, stability of concrete against water adsorption, stability of concrete against freezing, physical and mechanical properties and durabi
... Show MoreThe aim of this essay is to use a single-index model in developing and adjusting Fama-MacBeth. Penalized smoothing spline regression technique (SIMPLS) foresaw this adjustment. Two generalized cross-validation techniques, Generalized Cross Validation Grid (GGCV) and Generalized Cross Validation Fast (FGCV), anticipated the regular value of smoothing covered under this technique. Due to the two-steps nature of the Fama-MacBeth model, this estimation generated four estimates: SIMPLS(FGCV) - SIMPLS(FGCV), SIMPLS(FGCV) - SIM PLS(GGCV), SIMPLS(GGCV) - SIMPLS(FGCV), SIM PLS(GGCV) - SIM PLS(GGCV). Three-factor Fama-French model—market risk premium, size factor, value factor, and their implication for excess stock returns and portfolio return
... Show MoreWhen the drawdown pressure amounts to a value below the dew point pressure, a minor droplet of condensate is shaped and accumulated in the close area of wellbore. As the accumulation happens, the saturation of the liquid will grow and a reduction in gas relative permeability will happen, therefore it will affect the productivity. Generally, condensate baking problem in gas wells is being deliberated and studied and numerous techniques have been suggested to solve the problem. The studying of condensate banking dynamics is essential to evaluate the productivity and behavior of the wells of the gas fields.
Consider a simple graph on vertices and edges together with a total labeling . Then ρ is called total edge irregular labeling if there exists a one-to-one correspondence, say defined by for all where Also, the value is said to be the edge weight of . The total edge irregularity strength of the graph G is indicated by and is the least for which G admits edge irregular h-labeling. In this article, for some common graph families are examined. In addition, an open problem is solved affirmatively.
Photonic crystal fiber interferometers are used in many sensing applications. In this work, an in-reflection photonic crystal fiber (PCF) based on Mach-Zehnder (micro-holes collapsing) (MZ) interferometer, which exhibits high sensitivity to different volatile organic compounds (VOCs), without the needing of any permeable material. The interferometer is robust, compact, and consists of a stub photonic crystal fiber of large-mode area, photonic crystal fiber spliced to standard single mode fiber (SMF) (corning-28), this splicing occurs with optimized splice loss 0.19 dB In the splice regions the voids of the holey fiber are completely collapsed, which allows the excitation and recombination of core and cladding modes. The device reflection
... Show More