ABSTRACT
Machine learning (ML) has come a long way, as such has become a part of our everyday lives. Personal assistants like “Google Now” and “Apple Siri” to novel cybersecurity applications [1] use machine learning at their core. Deep learning, a form of machine learning, has enabled convolutional neural networks (CNNs) [2,3] to carry out classification of handwritten digits [4], complex face detection [5], self-driving cars [6], speech recognition [7], and much more. Coding by hand a problem such as speech recognition is nearly impossible because of the sheer amount of variance in the data. The way the human brain interprets these kinds of problems is complex, but they can be modeled to an extent using artificial neural networks (ANNs) with a substantial amount of accuracy, which sometimes even beat humans [8]. The whole concept of ANN started evolving after the 1950s. Artificial neurons gradually evolved from a simple perceptron to sigmoid neurons and then to many other forms. Earlier neurons were able to provide binary output to the many inputs that were provided. Newer algorithms and activation functions allow artificial neural networks to make complex predictions by learning on their own [2,3].
