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.
Background: Mental health is integrated into PHC as a strategy of WHO to fill the gap in mental health treatment. Part of this strategy needs a level of task shifting so that mental health care is provided by different level of PHC workers and not only specialists such as psychiatrists and psychologists.
Objectives: To assess the knowledge and attitudes of family doctors regarding the provision of psychological health in PHCC and if there is an association between the certificates of these family doctors and their Knowledge and attitudes to psychological health.
Subjects and Methods: A cross-sectional descriptive study with analytic elements was conducted in 8 famil
... Show MoreThis paper proposes a new approach, of Clustering Ultrasound images using the Hybrid Filter (CUHF) to determine the gender of the fetus in the early stages. The possible advantage of CUHF, a better result can be achieved when fuzzy c-mean FCM returns incorrect clusters. The proposed approach is conducted in two steps. Firstly, a preprocessing step to decrease the noise presented in ultrasound images by applying the filters: Local Binary Pattern (LBP), median, median and discrete wavelet (DWT), (median, DWT & LBP) and (median & Laplacian) ML. Secondly, implementing Fuzzy C-Mean (FCM) for clustering the resulted images from the first step. Amongst those filters, Median & Lap
This study is unique in this field. It represents a mix of three branches of technology: photometry, spectroscopy, and image processing. The work treats the image by treating each pixel in the image based on its color, where the color means a specific wavelength on the RGB line; therefore, any image will have many wavelengths from all its pixels. The results of the study are specific and identify the elements on the nucleus’s surface of a comet, not only the details but also their mapping on the nucleus. The work considered 12 elements in two comets (Temple 1 and 67P/Churyumoy-Gerasimenko). The elements have strong emission lines in the visible range, which were recognized by our MATLAB program in the treatment of the image. The percen
... Show MoreA solar updraft tower power plant (solar tower) is a solar thermal power plant that utilizes a combination of solar
air collector and central updraft tube to generate an induced convective flow which drives pressure staged turbines to generate electricity.
This paper presents practical results of a prototype of a solar chimney with thermal mass, where the glass surface is replaced by transparence plastic cover. The study focused on chimney's basements kind effect on collected air temperatures. Three basements were used: concrete, black concrete and black pebbles basements. The study was conducted in Baghdad from August to November 2009.
The results show that the best chimney efficiency attaine
... Show MoreEnhancing fatigue resistance in asphalt binders and mixtures is crucial for prolonging pavement lifespan and improving road performance. Recent advancements in nanotechnology have introduced various nanomaterials such as alumina (NA), carbon nanotubes (CNTs), and silica (NS) as potential asphalt modifiers. These materials possess unique properties that address challenges related to asphalt fatigue. However, their effectiveness depends on proper dispersion and mixing techniques. This review examines the mixing methods used for each nanomaterial to ensure uniform distribution within the asphalt matrix and maximize performance benefits. Recent research findings are synthesized to elucidate how these nanomaterials and their mixing proce
... Show MoreThis study was conducted to evaluate the efficacy of Saccharomyces cerevesiae as a growth promoting agent in tomato. Soaking the seeds in yeast suspension at 5 g/L for 12h increased germination percentage, root length, root fresh and dry weight, plant height, foliage fresh and dry weight, attained 88.5% ; 8.1 cm ; 84.3 mg ; 7.03 mg ; 10.75 cm ; 839 mg and 37.75 mg compared with 80% ; 5.33 cm ; 39 mg ; 4.8 mg ; 7.35 cm ; 608 mg and 25.5 mg in seedlings grown from non treated seeds respectively. Similar results were obtained with seedling from seeds soaked in S. cerevesiae filtrate for 12 hrs. with values of 77.5% ; 6.875 cm ; 91.5 mg ; 7.5 mg ; 9.5 cm ; 777 mg and 40.35 mg compared to 66% ; 5.8 cm ; 57.7 mg ; 5.03 mg ; 5.9 cm ; 493 mg
... Show MoreKinetics study on the phenol oxidation by catalytic wet air oxidation (CWAO) using CuO.NiO/Al2O3 as heterogeneous catalyst is presented. 4 g/l phenol solution of pH 7.3 was oxidized in a trickle bed reactor with gas flow rate of 80% stochiometric excess (S.E).. In order to verify the proposed kinetics, a series of CWAO experimental tests were done at two temperatures (140 and 160° C), oxygen partial pressures (9 and 12 bar), and weight hourly space velocity (WHSV) (1, 1.5, 2, 2.5, and 3 h-1). According to Power Law, the reaction orders are found to be approximately 1 and 0.5 with respect to phenol concentration and oxygen solubility, respectively. These values favorably compare with those cited in the literature for intrinsic kinetics,
... Show MoreUrban agriculture is one of the important urban uses of land in cities since the inception of cities and civilizations, but the great expansion of cities in the world during the twentieth century and the beginning of the twentieth century and the increase in the number of urban residents compared to the rural population has led to a decline in this use in favor of other uses.
This decline in agricultural and green land areas in cities has negatively affected the environment, natural life and biological diversity in cities in addition to the great impact on the climate and the increase in temperatures and the negative impact on the economic side, since urban agriculture is an important pillar of the economy, especially
... Show MoreSocial media is known as detectors platform that are used to measure the activities of the users in the real world. However, the huge and unfiltered feed of messages posted on social media trigger social warnings, particularly when these messages contain hate speech towards specific individual or community. The negative effect of these messages on individuals or the society at large is of great concern to governments and non-governmental organizations. Word clouds provide a simple and efficient means of visually transferring the most common words from text documents. This research aims to develop a word cloud model based on hateful words on online social media environment such as Google News. Several steps are involved including data acq
... Show More