Hello everyone, I am having some problems using Knime.
I am trying to build a deeplearner and testing it using the CIDDS-001 dataset.
The problem I am having is the DL4J Feedforward learner node only detects 5 columns which are (Duration, Src Pt, Dst Pt, Packets, AttackID).
I tried to use the domain classifier and remove the limit on the possible values, and made sure the data has no missing values. However, I am not
sure if Knime is detecting some values as missing values.
Any help would be appreciated.
I think you would have to convert the strings to numbers in an appropriate way.You could use OneHot Encoding but that might be to large a data set with possibly thousands of categories. Other methods could include converting to number (in the case of the Byte column with the M ? ^= thousands string included) or also the IP dresses might be used as numbers.
Only additional way would be to use the R package vtreat that usually does a very good job at providing a stable and useful conversion of strings to numbers without producing too large a dataset. I will see if I can come up with an example.
1 Like
Thank you @ maluber71.
Which kind of data i need to do a deeplearning model. Sorry I am new here and learning. When i run the model, knime did not recognize all columns for learning.