I'd like to know the cause of the following errors in the WEKA classifiers RandomForest and RandomTree:
ERROR RandomForest Execute failed: StackOverflowError in Weka during training. Please verify your settings.
ERROR RandomTree Execute failed: StackOverflowError in Weka during training. Please verify your settings.
The two nodes are contained within a loop that iterates through a list of files containing training and test data. The table structure is identical for each iteration. The two nodes run fine until a point towards the end of the loop when they throw these errors. I'm using the default settings for the nodes, but I have tried tweaking the settings to no avail. If anyone could give me some background on this error or some of the reasons it occurs, I would greatly appreciate it!
Some of the columns contain an abundance of zero values. However, I still receive these errors even after filtering out these columns. Also, I notice that when I shuffle the rows prior to partitioning, sometimes the nodes execute properly.
A sample workflow is attached. The two problematic nodes - RandomTree and RandomForest - are on the far right.