Solution A
A. k-NN based classifier
The first 3000 examples in the dataset have been used and value of k varies from 1 to 50.As you increase examples in the dataset,the percentage error decreases.
A graph with 'k' value on x-axis and 'percentage error on test set' on y-axis.
Solution B
B. Manifold based modeling of MNIST digits
Using Euclidean distance
Cluster for Digits 4 and 9
Cluster for Digits 1 and 7
Cluster for Digits from 0 to 9
As in the tenenbaum paper
Using Tangent distance
Cluster for Digits 4 and 9
Cluster for Digits 1 and 7
Cluster for Digits from 0 to 9
As in the tenenbaum paper
Solution C
C.Deep learning
Details