As one of the most ubiquitously applied unsupervised learning methods, clustering has also been known to have a few disadvantages. More specifically, parameters such as the number of clusters and neighborhood radius are usually unknown and hard to estimate in practical cases. Moreover, the stochastic nature of a great number of these algorithms is also a considerable point of weakness. In order to address these issues, we propose DISCERN which can serve as an initialization algorithm for K-Means, finding suitable centroids that increase the performance of K-Means. Following that, the algorithm can estimate the number of clusters if need be. The algorithm does all of that, while maintaining complete robustness and returning the same results at each separate run. We ran experiments on the proposed method processing multiple datasets and the results show its undeniable superiority in terms of results, computational time and robustness when compared to the randomized K-Means and K-Means++ initialization. In addition, the superiority in estimating the number of clusters is also discussed and we prove the lower complexity when compared to methods such as the elbow and silhouette methods in estimating the number of clusters.
442 Background: Previous reports suggested that colorectal cancers (CRC) appear at younger age in the United Arabs Emirates (UAE). However, these reports included nationals and expatriates in their analysis with heterogeneous population leading to biased analysis. our objective was to determine age and stage of disease in newly diagnosed UEA nationals with CRC treated at one major referral hospital in Abu Dhabi (SKMC) Methods: Charts of all patients diagnosed and/ or treated for CRC at SKMC between January 2000 and May 2011 were reviewed. Ultimately, only UEA nationals with diagnosis of adenocarcinoma of the colon and rectum were retained for further analysis. Results: Two hundreds six patients were diagnosed at SKMC as having colon or rectal carcinomas. Ninety two were expatriates and in 10 out of 113 nationals, the final diagnosis was squamous or adenosquamous carcinoma leaving 103 patients forming the population of this study. Median age was 57(10-100 years), with 59 men. The patient’s condition necessitated emergency operation in 38 and 15 either refused or were unfit for treatment mainly because of very advanced disease or severe co morbidity. The tumor location was: sigmoid: 46%, rectum: 33:, right colon :17% and 4% for descending and transverse colon. The stage of the disease was I; 4 pts, II: 16 pts, III 25 pts, IV in 49 pts and undetermined in 7. Resection was curative for 46 pts, palliative in 45, unknown in 10 and 2 refused surgery. Fifty one patients had their treatment essentially at SKMC and the others were treated both at SKMC and abroad. Average follow-up was 2 years and at last FU 39 were confirmed deceased, 29 were alive and the outcome in the remaining was unknown. Conclusions: UAE nationals with CRC presenting to our facility have relatively young age but presented with stage IV disease in half of the cases. Screening program for this population is warranted. Because a substantial number of patients had had their treatment abroad, rigorous follow-up and cancer outcome assessment was unreliable.
For mission critical (MC) applications such as bushfire emergency management systems (EMS), understanding the current situation as a disaster unfolds is critical to saving lives, infrastructure and the environment. Incident control-room operators manage complex information and systems, especially with the emergence of Big Data. They are increasingly making decisions supported by artificial intelligence (AI) and machine learning (ML) tools for data analysis, prediction and decision-making. As the volume, speed and complexity of information increases due to more frequent fire events, greater availability of myriad IoT sensors, smart devices, satellite data and burgeoning use of social media, the advances in AI and ML that help to manage Big Data and support decision-making are increasingly perceived as “Black Box”. This paper aims to scope the requirements for bushfire EMS to improve Big Data management and governance of AI/ML. An analysis of ModelOps technology, used increasingly in the commercial sector, is undertaken to determine what components might be fit-for-purpose. The result is a novel set of ModelOps features, EMS requirements and an EMS-ModelOps framework that resolves more than 75% of issues whilst being sufficiently generic to apply to other types of mission-critical applications.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.