A Machine learning Framework for Eye Melanoma Detection Based on SVM and CNN Classification

EasyChair Preprint 4263

Abstract

Eye skin cancer could be a rare malady however consistent with malignancy; it's the foremost common kind of cancer. similar to alternative varieties of cancers, it's curable for many of the cases if diagnosed properly however the method of diagnosing is sort of difficult and is that the most problematic issue within the treatment of eye skin cancer. This paper presents an automatic eye skin cancer detection technique employing a convolution neural network (CNN) and support vector machine (SVM) with victimization grey scale conversion for top image resolution. Two hundred pre-diagnosed samples square measure taken from a customary info followed by pre-processing to lower resolution samples and eventually fed to the CNN design. Though the projected technique needs a large computation, a high accuracy rate of ninety four.59% is achieved outperforming the attention willcer} detection victimization support vector machine classifier for feature classification and have extraction can implement the convolution neural network to extract options from the image.

Keyphrases: Convolution Neural Network, Pre-processing, eye melanoma, gray scale conversation, support vector machine classifier

@booklet{EasyChair:4263,
  author    = {Ankur Rao and Prashant Sharma and Ankita Bhargva},
  title     = {A Machine learning Framework for Eye Melanoma Detection Based on SVM and CNN Classification},
  howpublished = {EasyChair Preprint 4263},
  year      = {EasyChair, 2020}}