I would like to identify areas with statistically significant high occurrence versus areas of low occurrence using GIS analytic skills. I need to choose an analysis that is the most appropriate from below.

Hotspots analysis Proximity analysis Overlay analysis

I just developed a prediction model for a binary outcome. I graph the Area Under the ROC Curve (AUC) to investigate the __(which)___ of the model

Discrimination or Calibration

From the above question, the AUC of my prediction model is estimated as 0.5. I need the appropriate response.

This is a very good model in term of prediction performance.

The prediction performance is just moderate, not too good nor too bad.

This model is not useful.