How to i use optimal K in applying the model when i’m supposed to classify a market category?

ko2579

I’m supposed to split the dataset into 70% training and 30% test datasets. Apply the k-Nearest Neighbor algorithm using the optimal k. Interpret the results. Discuss the

performance of the predictive-type or classification-type learning. 

I tried using KNN clustering but i just ended up analysing the financial ratios. Please help🙏

Roland Jones_21245

Hi @ko2579,

Just to take each part of your question:

• You can split your data into a training/test split using either Split Data or Split Validation
• Make sure you use Set Role to set the correct variable to be your label
• To find the optimal k value, you could manually change the k value, or use the X Means operator which will do this for you. See this thread here: https://community.rapidminer.com/discussion/60056/clustering-k-means

Hope this helps.

Best,
Roland