Classification of a whole matrix

All,

I have a series of matrices containing signal returns from a set of objects under illumination. I wish to classify the objects in to 2 categories, A or B based on the contents of the signal return matrices.

In other words, I wish to classify whether a series of matrices are type A or B but using the data from the whole of each matrix to do this.

Is there a way to use the classification learner app to do this and if not, any pointers?

Thnanks.

Answer 1

[njj1]

You could turn the matrix into a vector (e.g., X = X(:)), then use the resulting vector as input to the classification learner. You need to have examples of A and B type matrices to design a classifier, or you could use a clustering algorithm to see if the matrices naturally cluster together.

Answer 2

[Image Analyst]

Well if we have no other information about the matrices, you could try deep learning. However if you know something about them, like they're images of either red flowers or blue flowers, then you could use traditional methods to measure those attributes, like is the matrix more red than blue, or more blue than red. Is there a spatial dependence among the elements? Why are the elements in a 2-D matrix rather than a 1-D vector? What distinguishes A from B? Can you simply look at the mean value? Or is it more complicated than that? You need to supply more context and show some small matrices as examples.

[Alex Gordon]

Apologies for being vague, the values in the matrices are RCS values for a range of targets as a function of angle to the antenna. We then hope to classify them as object type A or B.

I think the problem is more difficult to explain than I imagined so will give deep learning a look, many thanks.

[Image Analyst]

If you display them as images, do the A images look different somehow than the B images? Or do they both just look like random noise, or there is no discernible difference between the way they look?