I have a sparse dataset of size 775946 x 845372. I would like to perform a grid search in order to tune the parameters of my LogisticRegressionWithSGD model. I have noticed that the building of each model takes about 300 to 400 seconds. That means that in order to try all possible combinations of parameters I have to wait for about 24 hours. Most importantly though, I am not sure if the following combinations make sense at all. So, how should I pick up those parameters more wisely as well as in a way that I can wait less time?
val numIterations = Seq(100 , 500 , 1000 , 5000 , 10000 , 50000 , 100000 > , 500000) > val stepSizes = Seq(10 , 50 , 100 , 500 , 1000 , 5000 , 10000 , 50000) > val miniBatchFractions = Seq(1.0) > val updaters = Seq(new SimpleUpdater , new SquaredL2Updater , new > L1Updater) Any advice is appreciated. *// Adamantios*
