Hi Everyone, hope everybody is doing well in the pandemic.
I am currently using KNIME for my research work. My current is collecting the training data and then applying machine learning.
The data I have contains around 5.3 million rows and 1950 columns. Whenever, I start training the random forest, MLP Perceptron model it gives error like, “Java heap space”. Below the screen shot is attached.
I have tried again and again but the same problem occurs.
it looks like you need to increase your heap space for KNIME.
To do so, please have a look at this blog post:
The first bullet point in the section Optimize your knime.ini is what you are looking for.
If increasing your heapspace is not enough, I would start with less rows for the training.
Increase the number of rows slowly if the results are not sufficient.