search

lesliedb / spatial_cv

Repository for classwork
0 stars 2 forks source link

Directionally challenged #2

Open lesliedb opened 9 years ago

lesliedb commented 9 years ago

@dmcglinn I'm still a little unclear about the outputs from the k-fold method, but none the less I feel like I have a pretty firm grasp on the concept itself. Do you think I should focus on understanding the outputs or begin to start formulating how I will create the spatially auto-correlated data for the next part of the analysis?

dmcglinn commented 9 years ago

I suggest going on to tackle the autocorrelation component of the project as this is the most challenging aspect. Generalizing to k-folds is only necessary to provide computational efficiency.

dmcglinn commented 9 years ago

Definately take a look the Le Rest paper and the associated R code. https://sites.google.com/site/lerestk/publications That looks like the potentially most useful reference for you so far.