Assignment Description
Wine recognition data is given. Its number of features are:
class 1: 59, class 2: 71, class 3: 48.
NOTE: 1st dimension out of 15 identifies training and test split (1-training, 2-test). 2 nd dimension is class identifier (1-3).
Question 1. Distance Metrics Perform nearest neighbour classification experiments according to standard practices in pattern recognition. Use classification error as a fraction of incorrectly classified test points to compare different metrics from the course.
Question 2. K-means clustering Employ K-means to reduce the complexity of nearest neighbour classifier and compare the performance for different distance metrics.
Question 3. Neural Network Using Matlab Neural Network toolbox create a network, train and test with the wine data.
Attachment:- wine.data.rar