Please tell me the procedure how SMOreg technique works and how to analyse
As Regression technique is usually used for prediction , While using SMOreg
technique,it gives output in normalized form. I want to know how it can be
mapped predict the actual data..
Dear Weka Users,
in order to better understand the operating principles of parameter
optimization performed by GridSearch and CVParameterSelection I have had a
look at the code and I got a couple of doubts about the results produced,
in particular about the nested crossvalidations setting.
Both parameter search class and the classifier panel class perform
crossvalidation, therefore there is an "outer" crossvalidation, set by the
classifier panel class, that divides the whole dataset in n-fold subsets
and applies the parameter optimization to each of them collecting
fold-by-fold the predictions for all the instances, but there is also an
inner crossvalidation, set by the parameter search, that for each value of
the investigated parameter further splits each sub-dataset in m-fold
sub-sub-datasets, applies the classifier on each of them collecting the
predictions and finally returns the parameter best value (or parameters
values best combination).
This means that, most likely, each fold of the outer cv will correspond to
a classification model with different parameter(s) value(s), i.e. a
different model, therefore, whether I understood correctly, I am a bit
confused by the meaning of the predictions vector collected in this way.
The only explanation I deduced, is that this procedure does not aim to
validate the model but on the contrary the procedure itself.
Similar consideration can be done for the AttributeSelectedClassifer; also
in this case we have two nested crossvalidations and the model which the
different fold of the outer cv are performed with is likely different for
each fold. Therefore the summary statistics is calculated on a predictions
vector which different m-fold portions are likely related to m-fold
Thanks in advance for your precious help.
I have a dataset with 2320 instances and I performed a svm classification
with the optimized parameters c=500, gamma=0.1 . RMSE for training is
0.3122 which can be considered to be successful. Then, I wanted to supply a
test set (18 instances) to measure how well the model can predict the
correct results. The result was satisfactory with RMSE 0.3162, correctly
classified instances 15 and misclassified 3. The problem is when I wanted
the output predictions I saw that all the instances were misclassified.
There is a correlation with the misclassified values (ex. all excellent
values were predicted very bad, etc..) but none of them were predicted
well. I also supplied different data sets but the problem was the same. I
couldn't find the problem. How can I solve this issue? Output predictions
=== Predictions on test set ===
6,3:'VERY GOOD',1:VERY BAD,+,0.333
and the result window
=== Summary ===
Correctly Classified Instances 15 83.3333 %
Incorrectly Classified Instances 3 16.6667 %
Kappa statistic 0.7823
Mean absolute error 0.2259
Root mean squared error 0.3162
Relative absolute error 93.5738 %
Root relative squared error 85.7026 %
Coverage of cases (0.95 level) 100 %
Mean rel. region size (0.95 level) 83.3333 %
Total Number of Instances 18
I am using LibSVM classifier but it it runs very slow on my data (which is
somewhat big). So, I am wondering if there there is a chance to use
multithreading technique with LibSVM to speed up the classification process?
Thank you for your help,
View this message in context: http://weka.8497.n7.nabble.com/Multithreading-with-LibSVM-tp31344.html
Sent from the WEKA mailing list archive at Nabble.com.
I am using SMOreg in weka and I have to know how many support vectors were
used to make the model. Does weka give information about it? I can see the
model and count the support vectors from there, but the model is too long.
I am new to weka software and I am facing a problem in using grid search
method to optimize the parameters of the RBF kernel.
It shows the error message
"Problem evaluating classifier:"
"weka.classifiers.meta.gridsearch:cannot handle binary class !
Can anyone help me. I need it very urgently.
Dear all Wekalist members,
I have some questions about Random Forests
Algorithm as below:
1. Can we visualize forests as in C4.5 tree? In other words, in
C4.5 after classified the data we can easily visualize the tree (by R-click on
J48 classifier and the select visualize tree) but with random forest this
option not available is there any way to visualize random forests?
2. What is the type of random forests that is used in Weka? Where
in the main random forest algorithm that is proposed by (Leo Breiman 2001) he offered
different types of random forests called Forest-RI, Forest-RC each one has different
input selection (number of picked data to training it).
3. Also I am confusing in one more thing. I have dataset with more
than 200 features and thousands of instances once I used Random forests to
classify it working well and gave a good results but it selected only 10 random
features from 200 to construct trees and based on my knowledge random forests algorithm don’t use feature
selection. It selects sub features randomly from the main set of features to
construct each tree. So why it selected only 10 features and ignored the rest features
what are the criteria of selection?
4. Anyone has tried to use Random MultiNomial Logit in Weka? Which
is proposed by Anita Prinzie? Details in this article: Random Forests for
multiclass classiﬁcation: Random MultiNomial Logit.
Your help and
time is greatly appreciated.
Have a nice
day for all