In the previous chapter, we developed a project that accurately classifies cancer patients based on cancer types using an LSTM network. This is a challenging problem in biomedical informatics. Unfortunately, when it comes to classifying multimedia objects such as images, audio, or videos, linear ML models and other regular deep neural network (DNN) models, such as Multilayer Perceptron (MLP) or Deep Belief Networks (DBN), often fail to learn or model non-linear features from images.
On the other hand, convolutional neural networks (CNNs) can be utilized to overcome these limitations. In CNNs, the connectivity pattern between neurons is inspired by the human visual cortex, which more accurately resembles human vision, so it is perfect for image processing-related tasks. Consequently, CNNs have shown outstanding...
