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.
An experiment was conducted in a greenhouse - research station B - College of Agricultural Engineering Sciences, University of Baghdad, during the fall season of 2018 with the aim of propagating and initially studying the field performance of 18 and 20 potential potato lines derived from Rivera and Arizona cv. after in vitro exposure of nodal segments to different dosages of gamma rays (0, 10, 20, and 30 Gray) and EMS (0, 10, 20, and 30 mM). Each control cultivar and their derived lines were independently cultured in plastic bags according to the RCBD, with three replications. The results showed that the highest plant height and number of leaves were obtained from Arizona derived lines which gave 60.11 cm and 25.30 leaves.plant-1 in
... Show MoreAbstract
This study aims to identify the reality of using electronic applications in teaching language skills to people with mild intellectual disabilities from the mothers’ perspective. A descriptive approach was used. The electronic questionnaires were administered to the study sample, 122 responses were received from mothers of the students with mild intellectual disability in Hafer Al-Baten schools. The response average rate was 94%. The results showed that there are statistically significant differences that are related to the variant of monthly income as for the barriers to using electronic applications in such schools, whereas there were no differences regarding the variant of monthly income regarding t
... Show MoreA mathematical model constructed to study the combined effects of the concentration and the thermodiffusion on the nanoparticles of a Jeffrey fluid with a magnetic field effect the process of containing waves in a three-dimensional rectangular porous medium canal. Using the HPM to solve the nonlinear and coupled partial differential equations. Numerical results were obtained for temperature distribution, nanoparticles concentration, velocity, pressure rise, pressure gradient, friction force and stream function. Through the graphs, it was found that the velocity of fluid rises with the increase of a mean rate of volume flow and a magnetic parameter, while the velocity goes down with the increasing a Darcy number and lateral walls. Also, t
... Show MoreAbstract: Microfluidic devices present unique advantages for the development of efficient drug assay and screening. The microfluidic platforms might offer a more rapid and cost-effective alternative. Fluids are confined in devices that have a significant dimension on the micrometer scale. Due to this extreme confinement, the volumes used for drug assays are tiny (milliliters to femtoliters).
In this research, a microfluidic chip consists of micro-channels carved on substrate materials built by using Acrylic (Polymethyl Methacrylate, PMMA) chip was designed using a Carbon Dioxide (CO2) laser machine. The CO2 parameters have influence on the width, depth, roughness of the chip. In order to have regular
... Show More|
The research problem lies in the lack of accurate scientific perceptions concerning the reality of the communicator and the factors influencing his job. The research is aimed at introducing the communicator in the university press, clarifying the obstacles facing him, and uncovering the level of his job satisfaction and his visions of developing his work. The researcher adopted the survey method in the collection, analysis, and interpretation of the data using a questionnaire. A set of results and conclusions has been reached, most importantly are:
*The communicator performs multiple missions including writing, editing, and collecting info |
The current world is observing huge developments in presenting the opportunity for organizations and administrative units to use information and communication technology and their adoption by administrative work due to its importance in the achievement of work with higher efficiency, speed, and facility of communication with all individuals and companies using various means of communication Depending on the Internet networks. Therefore, the research dealt with the study of electronic systems designed and adopted in the creation or construction of a database for archiving data, which is the main method in organizations and administrative units in developed countries. Where this system works to convert documents, and manual processes and t
... Show MoreThe conservation of natural resources such as water is one of the areas that sustainable agriculture seeks to preserve, rationalize its use and protect it from pollution, because water is a specific factor for agriculture. Despite Iraq's possession of two international rivers, the Tigris and the Euphrates, which pour into the Shatt Al-Arab, it suffers from water shortages in recent times, As a result of several reasons combined, including water policies initiated by the riparian countries to reduce the quantities of water coming from the two rivers to Iraq, led by neighbor Turkey through the construction of dams and reservoirs and the establishment of irrigation projects, as well as climate changes from rising temperatures and lo
... Show MoreInformation hiding strategies have recently gained popularity in a variety of fields. Digital audio, video, and images are increasingly being labelled with distinct but undetectable marks that may contain a hidden copyright notice or serial number, or even directly help to prevent unauthorized duplication. This approach is extended to medical images by hiding secret information in them using the structure of a different file format. The hidden information may be related to the patient. In this paper, a method for hiding secret information in DICOM images is proposed based on Discrete Wavelet Transform (DWT). Firstly. segmented all slices of a 3D-image into a specific block size and collecting the host image depend on a generated key
... Show MoreIn this study, we review the ARIMA (p, d, q), the EWMA and the DLM (dynamic linear moodelling) procedures in brief in order to accomdate the ac(autocorrelation) structure of data .We consider the recursive estimation and prediction algorithms based on Bayes and KF (Kalman filtering) techniques for correlated observations.We investigate the effect on the MSE of these procedures and compare them using generated data.
In this paper, integrated quantum neural network (QNN), which is a class of feedforward
neural networks (FFNN’s), is performed through emerging quantum computing (QC) with artificial neural network(ANN) classifier. It is used in data classification technique, and here iris flower data is used as a classification signals. For this purpose independent component analysis (ICA) is used as a feature extraction technique after normalization of these signals, the architecture of (QNN’s) has inherently built in fuzzy, hidden units of these networks (QNN’s) to develop quantized representations of sample information provided by the training data set in various graded levels of certainty. Experimental results presented here show that
... Show More