Date created: | Last Updated:

Category: Project

Description: ABSTRACT One of the most prevalent cancers in the world is skin cancer, and prompt treatment and successful patient outcomes depend greatly on early identification and precise diagnosis. Skin cancer diagnosis, however, can be difficult because it calls for a dermatologist with extensive training and experience. Thus, the classification of skin cancer using deep learning models like CNNs has the potential to help in early identification and diagnosis of skin cancer. The HAM10000 dataset, which was used in this study, makes a substantial contribution to the field because it has a lot of excellent dermatoscopic images of different skin lesions. The suggested CNN model in this study is a deep learning model that performs exceptionally well in image classification tasks like the classification of skin cancer. It has numerous convolutional,pooling, and dense layers. The training data is oversampled, and the model's performance is enhanced by using the Adam optimizer to tweak its parameters. The model can learn the intricate patterns and characteristics present in the photos by being trained for 50 epochs with a batch size of 128. This improves classification performance. The best model is preserved when the Model Checkpoint callback is used, allowing for future use and reproducibility. Keywords: Machine learning, Convolutionl neural network, Dermatoscopic images, Conv2D, maxPooling 2D, Batch Normalization Cite this Article: Manjunath H R, Skin Cancer Detection Using Machine Learning, International Journal of Information Technology (IJIT), 4(2), 2023, pp. 10-19