Active learning (AL) has found wide applications in medical image segmentation, aiming to alleviate the annotation workload and enhance performance. Conventional uncertainty-based AL methods, such as entropy and Bayes...
详细信息
Active learning (AL) has found wide applications in medical image segmentation, aiming to alleviate the annotation workload and enhance performance. Conventional uncertainty-based AL methods, such as entropy and Bayes...
Active learning (AL) has found wide applications in medical image segmentation, aiming to alleviate the annotation workload and enhance performance. Conventional uncertainty-based AL methods, such as entropy and Bayesian, often rely on an aggregate of all pixel-level metrics. However, in imbalanced settings, these methods tend to neglect the significance of target regions, eg., lesions, and tumors. Moreover, uncertainty-based selection introduces redundancy. These factors lead to unsatisfactory performance, and in many cases, even underperform random sampling. To solve this problem, we introduce a novel approach called the Selective Uncertainty-based AL, avoiding the conventional practice of summing up the metrics of all pixels. Through a filtering process, our strategy prioritizes pixels within target areas and those near decision boundaries. This resolves the aforementioned disregard for target areas and redundancy. Our method showed substantial improvements across five different uncertainty-based methods and two distinct datasets, utilizing fewer labeled data to reach the supervised baseline and consistently achieving the highest overall performance. Our code is available at https://***/HelenMa9998/Selective_Uncertainty_AL.
There are many different sorts of data that can be gathered and analyzed, including pictures, videos, texts, speeches, music, and other noises, Video content, for example, generally includes minimum some types of audi...
详细信息
We present an overview of the second shared task on homophobia/transphobia Detection in social media comments. Given a comment, a system must predict whether or not it contains any form of homophobia/transphobia. The ...
详细信息
Heart disease prediction remains a critical area of research due to its significant impact on public health. This paper, titled “Improving Heart Disease Prediction with Stacked Ensemble Learning: A Comparison of Bina...
详细信息
ISBN:
(数字)9798331518943
ISBN:
(纸本)9798331518950
Heart disease prediction remains a critical area of research due to its significant impact on public health. This paper, titled “Improving Heart Disease Prediction with Stacked Ensemble Learning: A Comparison of Binary and Multiclass,” presents a comprehensive study comparing the performance of binary and multiclass classification models for heart disease prediction. Utilizing two distinct datasets, one for binary classification and one for multiclass classification. The study evaluates various machine learning and deep learning models, with a focus on stacked ensemble methods. The Binary Classification Results leveraging the Preprocessing-and-PCA-on-Heart-Disease-dataset, the study evaluated several models including Random Forest, Support Vector Machine (SVM), XGBoost, Gradient Boosting, Deep Learning model, and Logistic Regression. The results indicate that the Stacking Ensemble model achieved the highest accuracy of 90.2%, outperforming individual models where Random Forest (88.5%), SVM (89.1%), XGBoost (89.7%), Gradient Boosting (88.0%) and Deep Learning (89.1%). The Multiclass Classification study using the UCI Heart Disease dataset, when applied the same models, the Stacking Ensemble model achieved an accuracy of 64.1%, which, although superior to other individual models such as Random Forest (58.1%), SVM (54.3%), Gradient Boosting (59.2%) and Deep Learning model (57.6%) fell short compared to XGBoost, which attained an accuracy of 65.8%. This outcome highlights the challenges associated with multiclass classification in heart disease prediction.
Document-level Event Argument Extraction (DEAE) aims to identify arguments and their specific roles from an unstructured document. The advanced approaches on DEAE utilize prompt-based methods to guide pre-trained lang...
详细信息
Agriculture performs an critical position in India's economic system. Early detection of plant illnesses is critical to save you crop damage and similarly spread of diseases. Most plants, along with apple, tomato,...
Agriculture performs an critical position in India's economic system. Early detection of plant illnesses is critical to save you crop damage and similarly spread of diseases. Most plants, along with apple, tomato, cherry, grape, show symptoms of leaf ailment. These visible patterns may be found to correctly predict the disorder and take early movement to save you it. This can be triumph over with system getting to know and deep getting to know algorithms. We therefore recommend a method that determines tomato plant disease from pix of leaves. This method is performed with aid vector device (SVM), random woodland gadget studying algorithm, and look at algorithms Convolutional Neural Network (CNN), Recurrent Neural Network (RNN), and ResNet that is one of the switch learning techniques. Snoring. After the facts set is processed by way of the algorithms, the accuracy of the algorithms is compared and the snapshots are categorized.
This paper presents a novel decentralized architecture that aims to transform the metaverse by smoothly incorporating blockchain technology. The proposed framework promotes decentralization principles to address centr...
详细信息
ISBN:
(数字)9798350353839
ISBN:
(纸本)9798350353846
This paper presents a novel decentralized architecture that aims to transform the metaverse by smoothly incorporating blockchain technology. The proposed framework promotes decentralization principles to address centralization issues and restrict user empowerment in existing metaverse systems. It prioritizes putting people at the centre of their digital experiences. The decentralized framework goes beyond traditional tokenization methods by adding extra layers, like non-fungible tokens (NFTs) that change who owns an asset, and decentralized autonomous organizations (DAOs) that give people in communities more power, thereby fostering creativity and inclusion in virtual environments. Built upon the fundamental concepts of decentralization and user-centric design, the framework tackles the existing limits of the metaverse and envisions a future in which people actively define their virtual destiny. In addition to addressing issues, the framework provides possibilities for cooperative advancement, decentralized exchange of information, incorporation of future technology, and market growth, acting as a catalyst for innovation within the developing metaverse ecosystem. The framework embodies not only a framework but also a vision for a decentralized and user-centric metaverse, fostering creativity, inclusion, and the realization of virtual aspirations.
This paper summarizes the shared task on multimodal abusive language detection and sentiment analysis in Dravidian languages as part of the third Workshop on Speech and Language Technologies for Dravidian Languages at...
详细信息
In the era of big data, data trading significantly enhances data-driven technologies by facilitating data sharing. Despite the clear advantages often experienced by data users when incorporating multiple sources, the ...
详细信息
暂无评论