Use random forest classifier instead of svc.

Author: johann-petrak

.gitignore

@@ -24,4 +24,6 @@ hs_err_*.log
 !/tests/*/*/.keep
 !/tests/*/model-*/weka.yaml
 /test-tmp-EngineMalletSeqOut/
-
+/tests/cl-ionosphere/model-sklearn-external-rf/sklmodel*
+/tests/cl-ionosphere/model-sklearn-external-rf/indep.mtx
+/tests/cl-ionosphere/model-sklearn-external-rf/dep.mtx

tests/cl-ionosphere/train-sklearn-external-rf.xgapp

@@ -177,7 +177,7 @@
                 </entry>
                 <entry>
                   <string>algorithmParameters</string>
-                  <string>sklearn.svm.SVC</string>
+                  <string>sklearn.ensemble.RandomForestClassifier n_estimators 100 </string>
                 </entry>
                 <entry>
                   <string>featureSpecURL</string>