Neural_Networks

Neural Networks

Why Neural Networks at all?

There is no shortage of Machine Leanring Algorithms which rouses the question what advantages do Neural Networks offer. The different types of neural networks in deep learning, such as convolutional neural networks (CNN), recurrent neural networks (RNN), artificial neural networks (ANN) etc. are changing the way we interact with the world. These different types of neural networks are at the core of the deep learning revolution, powering applications like unmanned aerial vehicles, self-driving cars, speech recognition, etc. Even ML models can solve such problems but Deep Learning is preferred over ML by researchers and scientists as:

• Decision Boundary : For Classification problems non-linear decision boundaries cannot be trained on ML models.
• Feature Engineering : DL combines the feature extraction and feature selection and performs it automatically in its Hidden Layers for optimum performance.

Types of Neural Networks in Deep Learning:

There are mainly 3 important types of Neural Networks (NN) that form the foundations of most of the pre-trained models in DL:

• Convolutional Neural Network (CNN) : CNNs are currently booming and most used NN. CNNs are used in variety of Image and Video Processing Applications. Reader can find an excellent introductory article, Comprehensive Guide to CNNs

• Recurrent Neural Networks (RNN) : RNNs are the NNs with high potential and ever-growing set of applications and hard problems that need to be solve. RNNs are quite promising in terms of Research, both Academic and Industrial. RNNs are used to solve the problems related to the fields Text and Audio Data, i.e. Natural Language Processing and Audio Processing.

  Reader must go through the following articles for gaining fundamenta grasp on the topic:

  • Introduction to RNN, GFG - Basic introductory article to understand the use and scope of RNNs.
  • Recurrent Neural Networks at Depth, Medium - After digesting the first article, this article shall delve deeper into the concepts of RNN and the terms associated with it.
  • Recurrent Neural Networks Cheatsheet, Stanford - For the more experienced readers, this article shall provide for the some of the advanced know-hows in the field of NLP.

• Artificial Neural Networks (ANN) : A beginner article that would be best to get started with is, Guide to ANN, Medium

Sub Categories:

1. Deep Learning
2. Model SubClassing