The effect of spatio-temporal sample imbalance in epidemiologic surveillance using opportunistic samples: An ecological study using real and simulated self-reported COVID-19 symptom data.

Open surveys complementing surveillance programs often yield opportunistically sampled data characterised by spatio-temporal imbalance. We set up our study to understand to what extent spatio-temporal statistical models using such data achieve in describing epidemiological trends. We used self-reported symptomatic COVID-19 data from two Belgian regions, Flanders and the Brussels-Capital Region. These data were collected in a large-scale open survey with spatio-temporally imbalanced participation rates. We compared incidence estimates of both self-reported symptoms and test-confirmed COVID-19 cases obtained through generalised linear mixed models correcting for spatio-temporal correlation. We additionally simulated symptom incidences under different sampling strategies to explore the impact of sample imbalance, sample size and disease incidence, on trend detection. Our study shows that spatio-temporal sample imbalance generally does not lead to bad model performances in spatio-temporal trend estimation and high-risk area detection. Except for low-incidence diseases, collecting large samples will often be more essential than ensuring spatio-temporally sample balance.

Fault Diagnosis of Rolling Bearings in Primary Mine Fans under Sample Imbalance Conditions.

Rolling bearings are crucial parts of primary mine fans. In order to guarantee the safety of coal mine production, primary mine fans commonly work during regular operation and are immediately shut down for repair in case of failure. This causes the sample imbalance phenomenon in fault diagnosis (FD), i.e., there are many more normal state samples than faulty ones, seriously affecting the precision of FD. Therefore, the current study presents an FD approach for the rolling bearings of primary mine fans under sample imbalance conditions via symmetrized dot pattern (SDP) images, denoising diffusion probabilistic models (DDPMs), the image generation method, and a convolutional neural network (CNN). First, the 1D bearing vibration signal was transformed into an SDP image with significant characteristics, and the DDPM was employed to create a generated image with similar feature distributions to the real fault image of the minority class. Then, the generated images were supplemented into the imbalanced dataset for data augmentation to balance the minority class samples with the majority ones. Finally, a CNN was utilized as a fault diagnosis model to identify and detect the rolling bearings' operating conditions. In order to assess the efficiency of the presented method, experiments were performed using the regular rolling bearing dataset and primary mine fan rolling bearing data under actual operating situations. The experimental results indicate that the presented method can more efficiently fit the real image samples' feature distribution and generate image samples with higher similarity than other commonly used methods. Moreover, the diagnostic precision of the FD model can be effectively enhanced by gradually expanding and enhancing the unbalanced dataset.

[A survey of loss function of medical image segmentation algorithms].

Medical image segmentation based on deep learning has become a powerful tool in the field of medical image processing. Due to the special nature of medical images, image segmentation algorithms based on deep learning face problems such as sample imbalance, edge blur, false positive, false negative, etc. In view of these problems, researchers mostly improve the network structure, but rarely improve from the unstructured aspect. The loss function is an important part of the segmentation method based on deep learning. The improvement of the loss function can improve the segmentation effect of the network from the root, and the loss function is independent of the network structure, which can be used in various network models and segmentation tasks in plug and play. Starting from the difficulties in medical image segmentation, this paper first introduces the loss function and improvement strategies to solve the problems of sample imbalance, edge blur, false positive and false negative. Then the difficulties encountered in the improvement of the current loss function are analyzed. Finally, the future research directions are prospected. This paper provides a reference for the reasonable selection, improvement or innovation of loss function, and guides the direction for the follow-up research of loss function.

An end-to-end seed vigor prediction model for imbalanced samples using hyperspectral image.

Hyperspectral imaging is a key technology for non-destructive detection of seed vigor presently due to its capability to capture variations of optical properties in seeds. As the seed vigor data depends on the actual germination rate, it inevitably results in an imbalance between positive and negative samples. Additionally, hyperspectral image (HSI) suffers from feature redundancy and collinearity due to its inclusion of hundreds of wavelengths. It also creates a challenge to extract effective wavelength information in feature selection, however, which limits the ability of deep learning to extract features from HSI and accurately predict seed vigor. Accordingly, in this paper, we proposed a Focal-WAResNet network to predict seed vigor end-to-end, which improves the network performance and feature representation capability, and improves the accuracy of seed vigor prediction. Firstly, the focal loss function is utilized to adjust the loss weights of different sample categories to solve the problem of sample imbalance. Secondly, a WAResNet network is proposed to select characteristic wavelengths and predict seed vigor end-to-end, focusing on wavelengths with higher network weights, which enhance the ability of seed vigor prediction. To validate the effectiveness of this method, this study collected HSI of maize seeds for experimental verification, providing a reference for plant breeding. The experimental results demonstrate a significant improvement in classification performance compared to other state-of-the-art methods, with an accuracy up to 98.48% and an F1 score of 95.9%.

Fault diagnosis method of rolling bearing based on multiple classifier ensemble of the weighted and balanced distribution adaptation under limited sample imbalance.

Aiming at the minority samples cannot be effectively diagnosed when the samples are limited and imbalanced, a multiple classifier ensemble of the weighted and balanced distribution adaptation method (MC-W-BDA) is presented to solve the rolling bearing's fault diagnosis problem under the limited samples imbalance. We adopt random sampling to obtain enough different training sample sets whose base classifiers are trained in the Reproducing Kernel Hilbert Space. The appropriate base classifiers are integrated into strong classifiers by multiple classifier ensemble strategy to obtain the final result of classification. In addition, we propose A-distance method to automatically set the optimal parameter (balance factor) in MC-W-BDA. Experimental verification verifies the feasibility and effectiveness of proposed approach.

Predicting protein-RNA interaction amino acids using random forest based on submodularity subset selection.

Protein-RNA interaction plays a very crucial role in many biological processes, such as protein synthesis, transcription and post-transcription of gene expression and pathogenesis of disease. Especially RNAs always function through binding to proteins. Identification of binding interface region is especially useful for cellular pathways analysis and drug design. In this study, we proposed a novel approach for binding sites identification in proteins, which not only integrates local features and global features from protein sequence directly, but also constructed a balanced training dataset using sub-sampling based on submodularity subset selection. Firstly we extracted local features and global features from protein sequence, such as evolution information and molecule weight. Secondly, the number of non-interaction sites is much more than interaction sites, which leads to a sample imbalance problem, and hence biased machine learning model with preference to non-interaction sites. To better resolve this problem, instead of previous randomly sub-sampling over-represented non-interaction sites, a novel sampling approach based on submodularity subset selection was employed, which can select more representative data subset. Finally random forest were trained on optimally selected training subsets to predict interaction sites. Our result showed that our proposed method is very promising for predicting protein-RNA interaction residues, it achieved an accuracy of 0.863, which is better than other state-of-the-art methods. Furthermore, it also indicated the extracted global features have very strong discriminate ability for identifying interaction residues from random forest feature importance analysis.