Within the field of ‘Artificial Intelligence’ there exists ‘Machine Learning’, where we use computer algorithms to make predictions about previously unseen data (or situations), using learning (training) performed on existing datacomp.
Deep Learning is a subset of Machine Learning, and the technique that separates it from traditional Machine Learning is the focus on Neural Networks. These attempt to mimic organic systems, i.e. the brain, and the term ‘deep’ refers to the Neural Network having a ‘stack’ multiple layers. Deep Learning is considered a more modern technology than many others within Machine Learning.
Part of the reason why Deep Learning is a newcomer is that there is a heavy dependence on high end computer hardware. The huge number of multi-dimension calculations required for training Deep Learning models mean that only top specification computers can be relied on. With the minimum requirement only being feasible within the last twenty years. Traditional Machine Learning can very often be run on ‘normal’ business computers.
This huge amount of processing is a result of Deep Learning’s requirement for data; usually a great deal more in comparison to traditional Machine Learning. Traditional algorithms can be useful on a small amount of training data but the opposite is true for Deep Learning. number of Furthermore, whilst the requirement for data is high the workflow is also different, traditional Machine Learning is a series of discrete stages, Deep Learning is more of a single training process for
The difference between Deep and Traditional Machine Learning is seen in the actual training time