This study investigates how the hyperparameters epochs and batch size affect the classification accuracy of a convolutional neural network (CNN) trained on pulsar candidate data. Our results reveal that accuracy improves with increasing number of epochs and smaller batch sizes, suggesting that with optimized hyperparameters, high accuracy may be achievable with minimal training. These findings offer insights that could help create more efficient machine learning classification models for pulsar signal detection, with the potential of accelerating pulsar discovery and advancing astrophysical research.
The authors looked factors that could lead to earlier diagnosis of cardiovascular disease thereby improving patient outcomes. They found that advances in imaging and electrocardiography contribute to earlier detection of cardiovascular disease.
Levy & Levy tackle the optimization of the coronary artery bypass graft, a life-saving surgical technique that treats artery blockage due to coronary heart disease. The authors develop a novel auto-suturing method that saves time, allows for an increased number of sutures, and improves graft quality over hand suturing. The authors also show that increasing the number of sutures from four to five with their new method significantly improves graft quality. These promising findings may help improve outcomes for patients undergoing surgery to treat coronary heart disease.
Autonomous drone imaging combined with machine learning offers a promising approach for early detection of invasive species. In this study, students built an autonomous drone and compared three models: CNN, SGDC, and XGBoost, to identify Brassica nigra from aerial footage. Their results show that CNNs most effectively recognize key visual features, demonstrating strong potential for supporting conservation and invasive plant management.
The application of machine learning techniques has facilitated the automatic annotation of behavior in video sequences, offering a promising approach for ethological studies by reducing the manual effort required for annotating each video frame. Nevertheless, before solely relying on machine-generated annotations, it is essential to evaluate the accuracy of these annotations to ensure their reliability and applicability. While it is conventionally accepted that there cannot be a perfect annotation, the degree of error associated with machine-generated annotations should be commensurate with the error between different human annotators. We hypothesized that machine learning supervised with adequate human annotations would be able to accurately predict body parts from video sequences. Here, we conducted a comparative analysis of the quality of annotations generated by humans and machines for the body parts of sheep during treadmill walking. For human annotation, two annotators manually labeled six body parts of sheep in 300 frames. To generate machine annotations, we employed the state-of-the-art pose-estimating library, DeepLabCut, which was trained using the frames annotated by human annotators. As expected, the human annotations demonstrated high consistency between annotators. Notably, the machine learning algorithm also generated accurate predictions, with errors comparable to those between humans. We also observed that abnormal annotations with a high error could be revised by introducing Kalman Filtering, which interpolates the trajectory of body parts over the time series, enhancing robustness. Our results suggest that conventional transfer learning methods can generate behavior annotations as accurate as those made by humans, presenting great potential for further research.
Authors emphasize the challenges of manual tumor segmentation and the potential of deep learning models to enhance accuracy by automatically analyzing MRI scans.
In this study, the authors utilize an infrared camera to visualize and investigate the exothermic reaction of polyurethane foam, which has many everyday uses including automotive seats, bedding, and insulation.
