I have generated groundwater potential zones using AHP technique. Now I want to validate my result with existing yield data (77 wells). How can I know the sensitivity and specificity of my study?
If u want to create ROC curve inside ArcGIS then, you may use ArcSDM tool box. Further you can use R or SPSS also but u have to extract data to ASCII or other tabular format.
Thank you Moh. Debe and Achu Ashok for your answer. I have already use the SPSS to create ROC curve but I am not sure whether I have done right or wrong. when i am plotting prediction accuracy after separating the data set under Agree and Disagree it is coming 75% (by dividing Agreed value by total well yield data) but AUC is showing 65%. Before processing SPSS I have categorized my dataset into Agree with 1 and Disagree with 0. then i imported the actual yield data and Column marked with 0 and 1 to create ROC curve. Is this the right procedure to create ROC in SPSS? or if not please tell me proper procedure.
To generate ROC (Receiver Operating Characteristic) curves in the ArcGIS environment for validating groundwater potential zones, you can follow the steps below:
First, you need to have two datasets: the first one should be the groundwater potential zones that you have created, and the second one should be a dataset of the actual groundwater occurrence locations.
Convert both datasets into raster format using the 'Polygon to Raster' and 'Point to Raster' tools.
Use the 'Extract Multi Values to Points' tool to extract the values of the groundwater potential zones raster dataset at the location points of the groundwater occurrences raster dataset.
Create a new field in the groundwater occurrences attribute table and assign values of 1 to all the points.
Use the 'Merge' tool to merge the two raster datasets into a single raster dataset.
Use the 'Reclassify' tool to reclassify the merged raster dataset into two classes: the first one should represent the potential groundwater zones, and the second one should represent non-potential zones.
Use the 'Sample' tool to randomly sample a set of points from the reclassified raster dataset.
Use the 'Extract Multi Values to Points' tool to extract the values of the reclassified raster dataset at the sampled points.
Use the 'Summary Statistics' tool to calculate the true positive rate (TPR) and false positive rate (FPR) for each classification threshold.
Plot the TPR against the FPR to generate the ROC curve.
Calculate the area under the ROC curve (AUC) to evaluate the accuracy of the groundwater potential zones.
By following these steps, you can generate ROC curves in the ArcGIS environment for validating groundwater potential zones.