SimpleMKL.classification: Simple MKL In RMKL: Multiple Kernel Learning for Classification or Regression Problems

Description

This function conducts Simple MKL for precomputed gramm matrices

Usage

 ```1 2``` ```SimpleMKL.classification(k, outcome, penalty, tol = 10^(-4), max.iters = 1000) ```

Arguments

 `k` list of Gramm matrices `outcome` vector of binary outcome -1 and 1 `penalty` ppenalty of the smoothness of the resulting desicion rules `tol` change between to iterations is smaller than this, algorithms is considered to have converged `max.iters` maximum number of allowed iteratons

Value

gamma weight vector for the importnace of each kernel

alpha coeffiencents of the dual of MKL

time total amount of time to train model

max.iters Numvber of iterations to reach convergence criteria

Examples

 ``` 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17``` ```library(kernlab) library(caret) library(RMKL) #Load data data(benchmark.data) example.data=benchmark.data[[1]] # Split samples into training and test sets training.samples=sample(1:dim(example.data)[1],floor(0.7*dim(example.data)[1]),replace=FALSE) # Set up cost parameters and kernels C=100 kernels=rep('radial',3) degree=rep(0,3) scale=rep(0,3) sigma=c(0,2^seq(-3:0)) K=kernels.gen(example.data[,1:2], training.samples, kernels, degree, scale, sigma) K.train=K\$K.train SimpleMKL.classification(K.train,example.data[training.samples,3], C) ```

Example output

```Loading required package: lattice

Attaching package: ‘ggplot2’

The following object is masked from ‘package:kernlab’:

alpha

\$gamma
[1] 0 1 0

\$iters
[1] 2

\$alpha
[1] 2.145553e-03 4.408910e-05 3.305602e+01 9.999992e+01 1.510327e-04
[6] 1.006381e+01 5.955634e+01 1.367328e-04 7.490723e-05 9.357457e+01
[11] 3.009480e-05 9.999999e+01 3.355234e+01 4.510966e-05 4.157061e+01
[16] 4.322098e-05 2.878140e+01 3.517140e+01 5.006819e-05 8.655379e+00
[21] 7.704790e+01 3.096216e-03 4.199576e-04 3.663009e-04 9.485126e+01
[26] 2.022167e-01 1.063272e-05 3.830590e+01 5.403788e+01 4.769932e-05
[31] 4.550526e+01 2.488669e+01 4.254738e-05 9.330425e-05 1.000000e+02
[36] 1.938196e+01 7.059709e-06 2.879869e-04 1.000000e+02 1.881237e+01
[41] 9.999999e+01 7.290926e+01 1.368251e-05 1.833999e-03 2.513803e+00
[46] 1.215574e+01 6.457470e+01 1.224067e-04 1.723929e+01 1.491627e+01
[51] 1.192191e-05 4.693348e+01 8.235141e-05 9.999999e+01 9.195460e+00
[56] 2.384703e+01 1.000000e+02 3.568320e+01 3.981720e-05 6.119435e+01
[61] 1.099335e-04 9.335132e+01 3.283527e+01 1.831674e-05 5.352280e-05
[66] 3.922601e+01 2.661056e+01 9.999997e+01 1.136354e-05 1.297337e-02
[71] 3.788881e-05 2.106342e-05 1.745993e+01 8.288365e+01 1.675527e-05
[76] 8.198266e+01 8.287029e+01 1.229353e+01 5.506263e+01 2.461231e+01
[81] 5.173841e+00 2.725858e+01 8.698796e+01 2.131988e+01 1.363580e+01
[86] 3.905219e-05 3.075299e+01 1.188970e-05 7.514731e-04 1.984472e+01
[91] 3.463592e-05 1.579760e+01 9.631525e-06 1.191972e-04 1.361960e-04
[96] 1.528748e-05 7.926626e+01 3.415773e-04 8.772297e-04 1.000000e+02
[101] 3.373321e+01 1.232445e-04 9.999999e+01 9.306351e+01 4.843026e+01
[106] 3.344064e+01 6.260597e+01 4.263392e-05 3.717509e+01 1.000000e+02
[111] 1.962077e+01 7.166229e+01 6.048994e+00 2.966697e-05 3.720834e-05
[116] 3.955233e-04 2.736186e+01 1.848997e-05 3.661057e-05 6.439493e+01
[121] 2.902792e+01 8.906003e-06 9.999999e+01 3.642794e+01 1.666375e-04
[126] 7.206920e+01 4.882937e+01 1.023735e+01 2.965309e+01 3.079292e+01
[131] 1.740635e-05 2.116364e-04 9.999998e+01 3.207111e+01 2.579153e+01
[136] 8.814107e-06 6.054023e-05 1.096759e-04 8.122470e+01 4.045217e+01

\$b
[1] 0.03714253

\$gamma_all
\$gamma_all[[1]]
[1] 0.3333333 0.3333333 0.3333333

\$gamma_all[[2]]
[1] 0.0000000 0.5394908 0.4605092

\$gamma_all[[3]]
[1] 0 1 0
```

RMKL documentation built on May 2, 2019, 7:55 a.m.