I am performing a KNN classification for a sample with n=5 and m= 200 (n:# features, m:#samples). To find the optimum value for K, I calculated an error rate for each K (1:50). However, I found the the error rate vector as well as the classification_report are not fixed, and they are changing by each run.
This is it; the process of classification (through distance metrics) or regression (through majority voting) begins from an instance, generally, chosen randomly each time the algorithm is started.
As Mohammed Deeb said, you must specify a seed to start the k-NN.
It looks that the dataset has different shuffled each time. You need to fix the training and testing dataset , so the data must be the same for each time you change the value of k.
If you using MATLAB, please writ the following before your executing code in your m-file:
clear;
clear all;
It happens due to random variables.
- Badges
- Science topic
- Similar topics
- Data Mining
- Classification