1)k-NN based qualifier
(a)graph can be found at below given link
k vs %age error

(b)percentage error increases as value of k increase. For k=6 percentage error reaches minimum and then starts increasing again

2)Isomap:Isomap is one of several widely used low-dimensional embedding methods,
where geodesic distances on a weighted graph are incorporated with the classical scaling (metric multidimensional scaling).
Isomap is used for computing a quasi-isometric, low-dimensional embedding of a set of high-dimensional data points.
The algorithm provides a simple method for estimating the intrinsic geometry of a data manifold based on a rough estimate
of each data point's neighbors on the manifold.

(a) euclidean distance
i. 1-7 graph
ii. 4-9 graph
iii. All digits graph
(b) Tangent Distance
i. 1-7 graph
ii. 4-9 graph
iii. All digits graph

2D-isomap generated using tangent distance takes lot of time as compared to euclidean distance method.
(graph in case of tangent distance are for 500 test cases while in euclidean are for 3000 test cases)
Tangent distance isomap gives us more distinct clustering as compared to euclidean distance method.
Clearly in case of 4 and 9 clustering tangent distance is better. Also in alldigit clustering tangent distance is better.