machine learning - Similarity search between time series in Matlab. Possible ? I cant find R-tree implementation in matlab -

i implement similarity search in matlab. wanna know possible ?

my plan use 2 popular similarity measurement euclidean distance , dynamic time warping. both of these applied on time series dataset. question @ point how can evaluate both of these 2 measurement performance , accuracy ? seen literature saying should use k-nn algorithm.

then, plan apply dimensionality reduction on time series dataset. after reducued dimensionality of dataset. need index dataset using r-tree or indexing techniques available.

however problem this, need r-tree matlab code hardly able find in internet ...

i realised of implementation similarity search in c++, c , java ... im not familiar those. im hoping implement these in matlab ... guru me ?

also kind of evaluation can make evaluate performance each algorithm.

thanks

recently (r2010a believe), matlab added new functions k-nearest neighbor (knn) searching using kd-tree (a spatial indexing method similar r-tree) statistics toolbox. example:

load fisheriris                            % iris dataset q = [6 3 4 1 ; 5 4 3 2];                   % query points  % build kd-tree knnobj = createns(meas, 'nsmethod','kdtree', 'distance','euclidean');  % find k=5 nearest neighbors q [idx dist] = knnsearch(knnobj, q, 'k',5); 

refer this page nice description.

also if have image processing toolbox, contains (for long time now) implementation of kd-tree , knn searching. private functions though:

[matlabroot '\images\images\private\kdtree.m'] [matlabroot '\images\images\private\nnsearch.m'] 

to compare 2 approaches (dynamic time warping , euclidean distance), can design classic problem of classification; given set of labeled training/testing time series, task predict label of each test sequenceby finding similar ones using knn predict majority class. evaluate performance, use of standard measures of classification accuracy/error, etc..