topic K nearest neighbors or other conceptualization for Hot Spot Analyisis in Spatial Statistics Questions

K nearest neighbors or other conceptualization for Hot Spot Analyisis

HuaqiYuan — Thu, 21 Apr 2011 16:34:28 GMT

Hi there,

I have a dataset which are skewed (not normally distributed). And I have used SPSS to prove that by applying the normality test. Under this situation, do I use the K nearest neighbors with 8 neighbors as suggested by the documentation for the Hot Spot Analysis or other conceptualization like fixed distance band with neighbor parameter. I have found out the results from these two conceptualizations are quite different (maybe due to the skewed nature of the data). So I would like to know which conceptualization I should choose based on your experience.

Thanks!

Franky

Re: K nearest neighbors or other conceptualization for Hot Spot Analyisis

LaurenRosenshein — Tue, 14 Jun 2011 16:38:01 GMT

Hi Franky,

You are right that if your data is skewed you want to ensure that your features all have at least several neighbors, and 8 is a good rule of thumb. The question of whether to use K-nearest neighbors or fixed distance is really determined by the question that you're asking. Fixed distance is often a good option because it ensures that your scale is consistent across the whole study area, but if you want to ensure that all of your features have at least 8 neighbors what you might want to do is use the "Generate Spatial Weights Matrix tool", which allows you to set a fixed distance (and choose your fixed distance according to the question that you're asking), and then optionally lets you set a minimum number of neighbors. That way it will use the fixed distance band everywhere, but for those features where the fixed distance does not ensure that a feature has 8 neighbors, it will extend the distance just for those features to ensure they have the minimum number of neighbors that you set.

Hope this helps.

Lauren Rosenshein
Geoprocessing Product Engineer

Re: K nearest neighbors or other conceptualization for Hot Spot Analyisis

RyanWilliams1 — Tue, 11 Dec 2012 15:05:53 GMT

Hi Lauren,
Do you happen to know the algorithm behind the K nearest neighbor option? For example, is it measuring feature edge to feature edge or centroid to centroid?
Thanks. Ryan