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.
Aim:- to show that not all survival curves without shoulder are not able to repair or have lost the ability for the accumulation of sublethal damage.
Background:- the shoulder of the survival curve is considered as a
region of accumulation of sublethal damage also as an indicator for cell capacity to repair. The size can be influenced by the change ofthe slope of the linear portion of the survival curve.
Results:- we have shown that a survival curve with shoulder size of
1.5 Gy can be a straight line when the slope of the exponential part is changed so the
... Show MoreA fracture is a damage to bone tissue that causes damage to the tissue surrounding the bone and may penetrate the skin. Subjects and methods: the present study included (80) fractured Iraqi patients (and 40) patients with DM2 and (40) without DM2 and compared them with (40) healthy control. Patients and control are matched in age. This study showed a significant increase in retinol-binding protein 4 (RBP4) and a considerable decrease in Vit .A GPT and GOT in fracture patients with and without DM2. In addition, there was a significant negative correlation between RBP4 with (GPT and GOT) in fracture patients with DM2 and a significant positive correlation between RBP4 with (GPT and GOT) in fracture patients without DM2.
... Show MoreBackground: Recent research indicates that persistent inflammatory responses may contribute to the rise of diabetic nephropathy (DN) and diabetic cardiovascular disease (DCVD) in type 2 diabetes mellitus patients (DM2). Numerous molecules associated with inflammation and angiogenesis have been implicated in the development and progression of DN and DCVD, respectively. Methods: The subjects were separated into five groups: healthy controls (n= 25), type 2 diabetes mellitus patients (n= 30), type 2 diabetes mellitus patients with nephropathy DN (n= 30), and type 2 diabetes mellitus patients with cardiovascular disease DCVD (n= 30). The blood levels of irisin, IL-8, HbA1C, urea, and creatinine were determined. Results: In current study there w
... Show MoreThe Log-Logistic distribution is one of the important statistical distributions as it can be applied in many fields and biological experiments and other experiments, and its importance comes from the importance of determining the survival function of those experiments. The research will be summarized in making a comparison between the method of maximum likelihood and the method of least squares and the method of weighted least squares to estimate the parameters and survival function of the log-logistic distribution using the comparison criteria MSE, MAPE, IMSE, and this research was applied to real data for breast cancer patients. The results showed that the method of Maximum likelihood best in the case of estimating the paramete
... Show MoreThe fingerprints are the more utilized biometric feature for person identification and verification. The fingerprint is easy to understand compare to another existing biometric type such as voice, face. It is capable to create a very high recognition rate for human recognition. In this paper the geometric rotation transform is applied on fingerprint image to obtain a new level of features to represent the finger characteristics and to use for personal identification; the local features are used for their ability to reflect the statistical behavior of fingerprint variation at fingerprint image. The proposed fingerprint system contains three main stages, they are: (i) preprocessing, (ii) feature extraction, and (iii) matching. The preprocessi
... Show More