• Heta Desai

In recent years, deep artificial neural networks, including recurrent ones, have achieved significant success in various pattern recognition and machine learning challenges. This overview summarizes key contributions, many of which date back to the previous century. The distinction between shallow and deep learners lies in the depth of their credit assignment paths, which are sequences of potentially learnable, causal connections between actions and outcomes. The review covers deep supervised learning (including a recap of the backpropagation history), unsupervised learning, reinforcement learning, evolutionary computation, and indirect methods for discovering short programs that represent deep, large networks.

A Comprehensive Review of Deep Learning Techniques in Neural Networks

"A Comprehensive Review of Deep Learning Techniques in Neural Networks", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.5, Issue 1, page no.400-413, January-2018, Available :http://www.jetir.org/papers/JETIR1801375.pdf

"A Comprehensive Review of Deep Learning Techniques in Neural Networks", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.5, Issue 1, page no. pp400-413, January-2018, Available at : http://www.jetir.org/papers/JETIR1801375.pdf