harry, It all depends on your the dataset, the parameters used and the desired output. I’m not currently running any test suite to give you specific parameters, but even so, the parameters should be adjusted based on your data. There are no ‘magic’ values that works in all cases. Sure, the DBSCAN can be a great option to detect outliers but I’m not sure that it was designed to do specifically this task. I have just mentioned one example where DBSCAN could be used, but it can also be used in a bunch of other scenarios.