1Assistant Professor, Department of CSE, S R Engineering College, Warangal
2Senior Assistant Professor, Research Scholor, Osmania University
3Research Scholor Associate Professor, Department of CSE, S R Engineering College, SR University, Warangal
4Assistant Professor, Department of CSE, Sumathi Reddy Institute of Technology and Science, Warangal, India
Online published on 7 December, 2018.
This paper investigates the opportunities for applying deep learning networks to tumor classification. It finds that basic networks can be found to convey sensible performance, equivalent with mid-run entertainers on the same dataset. Model saturation is a significant issue which can be settled by a combination of limiting the quantity of parameters in the model, include ensuring that training data is adjusted amongst positive and negative perceptions, low learning rates, and iteratively biasing the input data towards cases that the model has mis-arranged after past training epochs.
Deep learning networks, CNN, tumour classification
