Deep Learning

Deep learning is a branch of machine learning that utilizes models called artificial neural networks, the architecture of which have been inspired by human neurons. & Although artificial neural networks have been available for quite some time, deep learning or training deeper networks have recently become possible due to the availability of faster computers and larger data sets. With many trainable subsequent layers, deep networks can be trained in an end-to-end manner and learn hierarchical features of the input. Deep networks can provide state-of-the-art results for various problems such as supervised, unsupervised, or reinforcement learning. Recently, there have been many variations of networks in terms of architectures, layers, and loss functions. Still, they all share the same principle, which is to utilize differentiable layers in order to train networks parameters effectively with back-propagation.