9ba6a5a0

This book provides a good start point for Deep learning in R. It starts by defining deep learning and introducing related packages in R. Specifically it talks about H2O package an interface to the H2O software a powerful library for deep learning and machine learning. Next it shows how we can apply neural network algorithm on particular problem with one chapter about different approaches to avoid of overfitting data. It continues with dealing with anomalous data and how unsupervised learning and auto-encoder model can be used to identify atypical data. In the other chapter it elaborates supervised learning, with focus on training and building feedforward neural networks to develop prediction models. The book is ended with two approaches for optimizing models i.e. dealing with missing data and choosing best hyper parameters value.