Extreme Learning Machine for Breast Cancer Diagnosis using cloud computing
Priya Shirley Muller, Mudarakola Lakshmi Prasad, Poornima A Sikrant, Pundru Chandra Shaker Reddy, Swati Sharma, Nipun Sharma · 2024
Worldwide, breast cancer ranks high among women's leading causes of mortality. The likelihood of mortality from breast cancer can be decreased with early detection and rapid treatment. Among those residing in rural regions with limited access to healthcare, machine learning on the cloud is currently playing a crucial role in illness diagnosis. Machine learning-based diagnostic systems can supplement radiologists' work as secondary readers to ensure accurate illness diagnosis, while cloud-based systems can facilitate remote diagnostics and telehealth services. The potential of artificial neural networks (ANN)-based techniques for illness diagnosis has piqued the interest of numerous academics. Among the many ANN versions, extreme learning machine (ELM) stands out as a promising candidate for tackling a wide range of classification issues. Three distinct areas of study are brought together in the suggested framework of this work: The first area where ELM is used is in the detection of breast-cancer. Next, the gain-ratio feature selection strategy is utilized to remove features that are not significant. Finally, this research offers an ELM-based CC-approach for remote breast-cancer diagnostics. Several cutting-edge technologies for illness diagnosis are contrasted with the cloud-based ELM's performance. The outcomes obtained on the Wisconsin-Diagnostic-Breast-Cancer(WBCD) dataset demonstrate that the ELM technique hosted in the cloud surpasses alternative approaches. When comparing ELM's performance in both the standalone and cloud contexts, the former proved to be superior.