Merge pull request #1914 from Saurabh7/multiclassnb

extend multiclass nb

doc/ipython-notebooks/multiclass/multiclass_reduction.ipynb

@@ -142,7 +142,7 @@
       "Xall = mat['data']\n",
       "\n",
       "#normalize examples to have norm one\n",
-      "Xall = Xall / sqrt(sum(Xall**2,0))\n",
+      "Xall = Xall / np.sqrt(sum(Xall**2,0))\n",
       "Yall = mat['label'].squeeze()\n",
       "\n",
       "# map from 1..10 to 0..9, since shogun\n",
@@ -458,6 +458,64 @@
      "language": "python",
      "metadata": {},
      "outputs": []
+    },
+    {
+     "cell_type": "heading",
+     "level": 3,
+     "metadata": {},
+     "source": [
+      "Using a kernel multiclass machine"
+     ]
+    },
+    {
+     "cell_type": "markdown",
+     "metadata": {},
+     "source": [
+      "Expanding on the idea of creating a generic multiclass machine and then assigning a particular multiclass strategy and a base binary machine, one can also use the [KernelMulticlassMachine](http://www.shogun-toolbox.org/doc/en/current/classshogun_1_1CKernelMulticlassMachine.html) with a kernel of choice.\n",
+      "\n",
+      "Here we will use a [GaussianKernel](http://www.shogun-toolbox.org/doc/en/3.0.0/classshogun_1_1CGaussianKernel.html) with LibSVM as the classifer.\n",
+      "All we have to do is define a new helper evaluate function with the features defined as in the above examples."
+     ]
+    },
+    {
+     "cell_type": "code",
+     "collapsed": false,
+     "input": [
+      "def evaluate_multiclass_kernel(strategy):\n",
+      "    from modshogun import KernelMulticlassMachine, LibSVM, GaussianKernel\n",
+      "    width=2.1\n",
+      "    epsilon=1e-5\n",
+      "    \n",
+      "    kernel=GaussianKernel(feats_train, feats_train, width)\n",
+      "    \n",
+      "    classifier = LibSVM()\n",
+      "    classifier.set_epsilon(epsilon)\n",
+      "\n",
+      "    mc_machine = KernelMulticlassMachine(strategy, kernel, classifier, lab_train)\n",
+      "\n",
+      "    t_begin = time.clock()\n",
+      "    mc_machine.train()\n",
+      "    t_train = time.clock() - t_begin\n",
+      "\n",
+      "    t_begin = time.clock()\n",
+      "    pred_test = mc_machine.apply_multiclass(feats_test)\n",
+      "    t_test = time.clock() - t_begin\n",
+      "\n",
+      "    evaluator = MulticlassAccuracy()\n",
+      "    acc = evaluator.evaluate(pred_test, lab_test)\n",
+      "\n",
+      "    print \"training time: %.4f\" % t_train\n",
+      "    print \"testing time:  %.4f\" % t_test\n",
+      "    print \"accuracy:      %.4f\" % acc\n",
+      "\n",
+      "print \"\\nOne-vs-Rest\"\n",
+      "print \"=\"*60\n",
+      "evaluate_multiclass_kernel(MulticlassOneVsRestStrategy())\n",
+      "\n"
+     ],
+     "language": "python",
+     "metadata": {},
+     "outputs": []
     }
    ],
    "metadata": {}