Accelerated Asynchronous Greedy Coordinate Descent Algorithm for Kernel SVMs (AsyAGCD) on multicore shared memory system

AsyAGCD is based on the parallel asynchronous greedy coordinate descent algorithm (Asyn-KSVM) which proposed by Yang You et. al.

For more details about this algorithm please refer to the following papers:

Accelerated Asynchronous Greedy Coordinate Descent Algorithm for Kernel SVMs (AsyAGCD) 

Build

We require the following environment to build AsyAGCD:

• Unix Systems (We haven't tested on Mac OS).

To build the program, simply run make. Two binaries, svm-train (for training) and svm-predict (for prediction) will be built.

Data Preparation

Please download the datasets from LIBSVM datasets http://www.csie.ntu.edu.tw/~cjlin/libsvmtools/datasets

Usage

./svm-train is used for parallel training, and it can print out the prediction accuracy (for C-SVC) and MSE or MAE (for \epsilon-SVR) at each step.

Usage: svm-train [options] training_set_file testing_set_file [model_file]
options:
options:
-s svm_type : set type of SVM (default 0)
    0 -- C-SVC        (classification)
    3-- epsilon-SVR        (regression)
-t kernel_type : set type of kernel function (default 2)
    0 -- linear: u'*v
    1 -- polynomial: (gamma*u'*v + coef0)^degree
    2 -- radial basis function: exp(-gamma*|u-v|^2)
    3 -- sigmoid: tanh(gamma*u'*v + coef0)
-d degree : set degree in kernel function (default 3)
-g gamma : set gamma in kernel function (default 1/num_features)
-r coef0 : set coef0 in kernel function (default 0)
-c cost : set the parameter C of C-SVC, epsilon-SVR, and nu-SVR (default 1)
-p epsilon : set the epsilon in loss function of epsilon-SVR (default 0.1)
-m cachesize : set total cache memory size in MB (default 2000); each thread will use cachesize/n memory
-e epsilon : set tolerance of termination criterion (default 0.001)
-h shrinking : whether to use the shrinking heuristics, 0 or 1 (default 1)
-wi weight : set the parameter C of class i to weight*C, for C-SVC (default 1)
-v n: n-fold cross validation mode
-T maxiter : Maximum number of iterations (default 10*(number of samples))
-I i : print accuracy after every i iterations (default maxiter/10)
-N nthreads : number of threads
-u replication : number of sum-of-margins SVOR replication (default 2)

Note that the prediction accuracy printed out at each iteration is for each 1-vs-1 subproblem for multi-class classification.

./svm-predict is used for prediction:

Usage: svm-predict [options] test_file model_file output_file
options:
-b probability_estimates: whether to predict probability estimates, 0 or 1 (default 0)
-q : quiet mode (no outputs)

Additional Information

If you have any questions or comments, please send an email to [email protected]. We appreciate your feedback.