classification

I am representing each XML document as a feature matrix in a csr_matrix format. Now that I have around 3000 XML documents, I got a list of csr_matrices. I want to flatten each of these matrices to become feature vectors, then I want to combine all of these feature vectors to form one csr_matrix representing all the XML documents as one, where each row is a document and each column is a feature. One way to achieve this is through this code X= csr_matrix([a.toarray().ravel().tolist() for a in ls]) where ls is the list of csr_matrices, however, this is highly inefficient, as with 3000 documents,

I'm trying to make my classification process a bit faster. I thought of increasing the first input_dim in my deploy.prototxt but that does not seem to work. It's even a little bit slower than classifying each image one by one. deploy.prototxt input: "data" input_dim: 128 input_dim: 1 input_dim: 120 input_dim: 160 ... net description ... python net initialization net=caffe.Net( 'deploy.prototxt', 'model.caffemodel', caffe.TEST) net.blobs['data'].reshape(128, 1, 120, 160) transformer = caffe.io.Transformer({'data':net.blobs['data'].data.shape}) #transformer settings python classification images=

I'm facing a text classification problem, and I need to classify examples to 34 groups. The problem is, the size of training data of 34 groups are not balanced. For some groups I have 2000+ examples, while for some I only have 100+ examples. For some small groups, the classification accuracy is quite high. I guess those groups may have specific key words to recognize and classify. While for some, the accuracy is low, and the prediction always goes to large groups. I want to know how to deal with the "low frequency example problem". Would simply copy and duplicate the small group data work? Or

I would like to use 10-fold Cross-validation to evaluate a discretization in MATLAB. I should first consider the attributes and the class column. In Statistics Toolbox there is CROSSVAL function, which performs 10-fold cross validation by default. Check it out. Another function CROSSVALIND exists in Bioinformatics Toolbox. Also there is an open source Generic-CV tool: http://www.cs.technion.ac.il/~ronbeg/gcv/ If you would rather write your own xval wrapper rather than using built-in functions, I often use randperm() to generate random orderings of my data, which you can then partition using a