Three-dimensional array

Question

In a classification problem, I have c classes of data, each with e examples of this data, and each example is represented by a feature vector of length

Answer 1

The restriction with a 3D matrix is that dimensions must have a fixed length (so all classes would have to have the same number of examples, and features).

Assuming that you want a general answer, and providing you want to avoid cell arrays, you could use a structure array:

c(3).name = 'cls3';  %// name of 3rd class

c(3).example(1).name = 'c3e1';    %// name of 1st example of 3rd class 
c(3).example(1).data = [1 2 3 4];  %// not sure what is your data structure here
c(3).example(1).features_ID = [101 102];  %// ID of features

c(3).example(2).name = 'c3e2';
c(3).example(2).data = [1 2 3 4 5 6];
c(3).example(2).features_ID = [101 102 103];

I only filled in the 3rd class, with 2 examples, but you get the idea.

Answer 2

Generally it's best to place the longest vector in a column. Therefore, (f, e, c) should be better than (c, e, f), assuming that f is indeed the longest dimension.

MATLAB is most efficient when working in columns, and many built-in functions are coded to work columnwise by default.

Source: http://www.mathworks.com/help/images/using-columnwise-processing-to-speed-up-sliding-neighborhood-or-distinct-block-operations.html.