I am reading the literature on sequential learning and it is often mention that in order to handle sequential/temporal data, there two categories of work in sequential learning
- External Representation of Time
- Internal Representation of Time
I am not able to understand the difference between two, for example in the paragraphs below, it is hinted that external representation can be inserted directly by
preprocessing time through a time to space transformation. Following example is given to elaborate this.
For instance, one effective way of constructing a spatial representation of temporally- occurring information (which is not only limited to neural computing) is to create the power spectrum of the incoming information and use it as a static input image.
Is there any simple and intuitive example to understand this? I cant get time to space transformation or power spectrum thing.