CrazyEngineers
CrazyEngineers

  • k nn classifier

    sherly

    Member

    Updated: Oct 26, 2024
    Views: 1.5K
    hi
    i am working in classification of texture images.
    can anyone help me with a code for k nearest neighbour classifier and also help me to create a training dataset with labelled categories.
    0
    Replies
Howdy guest!
Dear guest, you must be logged-in to participate on CrazyEngineers. We would love to have you as a member of our community. Consider creating an account or login.
Replies

  • deepmali

    Member • Feb 16, 2009

    hi
    I am also doing something similar..I am using SVD based approach...If you have any info pls post it here.
    Are you sure? This action cannot be undone.
    Cancel

  • sherly

    Member • Feb 23, 2009

    classification rate

    hi,
    I am working on image classification.
    I am using minimum distance classifier.
    Can anyone help me with the formula for correct classification rate.
    Thank you.
    Are you sure? This action cannot be undone.
    Cancel

  • Predictor

    Member • Mar 14, 2009

    Re: classification rate

    sherly
    hi,
    I am working on image classification.
    I am using minimum distance classifier.
    Can anyone help me with the formula for correct classification rate.
    I think that what you are referring to is simply the accuracy of the classifier, which is defined as: the number of correct classifications divided by the total number of classifications.


    -Will Dwinnell
    <a href="https://matlabdatamining.blogspot.com/" target="_blank" rel="nofollow noopener noreferrer">Data Mining in MATLAB</a>
    Are you sure? This action cannot be undone.
    Cancel

  • Predictor

    Member • Mar 14, 2009

    sherly
    hi
    i am working in classification of texture images.
    can anyone help me with a code for k nearest neighbour classifier and also help me to create a training dataset with labelled categories.
    k-nearest neighbors is very simple:

    1. Calculate the distance from the current case to all historical cases.
    2. Select the k historical cases with the least distance from the current case.
    3. Merge those k cases, typically by averaging for numeric estimation problems or by voting for classification problems.


    -Will Dwinnell
    <a href="https://matlabdatamining.blogspot.com/" target="_blank" rel="nofollow noopener noreferrer">Data Mining in MATLAB</a>
    Are you sure? This action cannot be undone.
    Cancel
Home Channels Search Login Register