A review on Neural Turing Machine
This is an incremental review paper summarizing existing research on NTMs for the AI community.
The paper conducts a systematic review of Neural Turing Machines (NTMs), analyzing their concepts, structure, versions, tasks, and comparisons, and concludes with discussions on issues and future work.
One of the major objectives of Artificial Intelligence is to design learning algorithms that are executed on a general purposes computational machines such as human brain. Neural Turing Machine (NTM) is a step towards realizing such a computational machine. The attempt is made here to run a systematic review on Neural Turing Machine. First, the mind-map and taxonomy of machine learning, neural networks, and Turing machine are introduced. Next, NTM is inspected in terms of concepts, structure, variety of versions, implemented tasks, comparisons, etc. Finally, the paper discusses on issues and ends up with several future works.