Hello. I am very new to KNIME Analytics and Data Analytics. I was working on this project to actually do win prediction for me. I have making a datasets of about 100 data but the scorer at the end usually just show prediction for around 20 data. Am I making a mistakes during making the model?
My Model basically
File Read → Partition → Random Forest Learner → Random Forest Predictor-> Scorer (Javascript)
Any answer regarding scorer and confusion matrix is really appreciated. Thanks a lot!
@darrianjovan welcome to the KNIME forum. Is it possible the scored data you see is from the test data? You might want to take a look at some ressources to know more about data mining and predictive models:
Hi. Thanks for the recommendation. I’ll check on it later on. Here I uploaded the raw data I used for the model and the picture of the process.
How can I change the scorer from test data instead of the training?
Raw Data.xlsx (9.3 KB)
@darrianjovan the grey square would hold the model which has been trained on the portion of the data split by the partition node to the top. And then been tested on the part on the bottom, which is the fraction of the data you noticed.
You can use the model (save it) to score any data that would have the same structure.
From your questions I would assume you might benefit a great deal from some basic introductions into machine learning.
This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.
