Deep Learning for Time-Series

Deep learning is a subfield of machine learning concerned with algorithms relating to neural networks. Neural networks, or, more precisely, artificial neural networks (ANNs) got their name because of the loose association with biological neural networks in the human brain.

In recent years, deep learning has been enhancing the state of the art across the bench in many application domains. This is true for unstructured datasets such as text, images, video, and audio; however, tabular datasets and time-series have so far shown themselves to be less amenable to deep learning.

Deep learning brings a very high level of flexibility and can offer advantages of both online learning, as discussed in Chapter 8, Online Learning for Time-Series, and probabilistic approaches, as discussed in Chapter 9, Probabilistic Models for Time-Series. However, with its highly parameterized models, finding the right model can be a challenge.

Among the contributions...