User:Passarel/sandbox

In order to allow the analysis of temporal data in a sensor network, several aspects need to be considered. One of the main factors regards the temporal nature of the data, and the sparsity of the measurements. Also, it is important to allow some exploratory data analysis, so the experts can have a better understanding of the domain. In this wiki, some general guidelines will be given to perform this kind of analysis, as well as example of code to perform the task and examples.

Exploratory data analysis

[Graphical visualizations]

[Statistical metrics]

[Principal Components Analysis]

Grouping temporal intervals

[Criteria to define atomic intervals]

[Basic statistics for sparse measurements]

[Shapelets]

Learning in temporal data

[Defining intervals]

[Sliding Windows]

[Recurrent neural networks]

[Convolutional networks for temporal learning]