Data & Analytics Exam > Data & Analytics Videos > Weka Tutorial > Weka Tutorial 38: Learning Curves 2 (Model Evaluation)
Weka Tutorial 38 Learning Curves 2 (Model Evaluation) Video Lecture - Weka
Text Transcript from Video
our YouTube welcome to another wake a
tutorial this tutorial is the
continuation of tutorial 36 where I
demonstrated how to plot learning curves
for different percentage of data size
now in this case also learning curves
can be produced for different parameters
or the change of the parameters of your
learning algorithms so I promised in
Wicca tutorial 36 that I'll be back with
these version of producing learning
curve by changing parameters of your
learning algorithm this is how you do it
so we have our way cut here and we are
going to click experimenter we are
working with experimenter and in the
experimenter we have an option
configuration mode option to set to
advanced we click on new and in the
destination it's a good practice to let
click on this particular example and
option you see that if I expand that
then we have an output file so this is
just to get output file the results
you're producing from Weka
so I'm just setting up to desktop file
called one go dark file so here all of
my results will be published so this is
a minor thing again it's it's not
necessary to understand what's going on
here it's just to save your results in
some output file so you click on ok so
that's the destination part next comes
result generator this is the important
thing to understand if you're not
following this tutorial and you can go
back to tutorial 36 the links approve
links link is provided here with the
video now in the result generator you
have the option you have some options to
choose from we're choosing the second
option cross-validation result producer
and then we are clicking left clicking
on the option here now here again you
have a bunch of options coming up number
of false output file raw code but
not interested in them were interested
in split evaluator so in this big
evaluator you can see that you have
again options click on those options and
something pops up and you see that we
have a bunch of options for the generic
object editor but we are not interested
any of them except for the classifier
from the classifier you choose let me
take it to the top you choose a meta
algorithm called filtered search you
select that and again you are left
clicking on the option here and now we
are going to set the classifier of our
of our interest which is going to be
random forests in this case and in the
end you just click OK to this window ok
to this window and ok to this window now
we have setup the result generator now
we have a generator properties here
which is disabled by default you're
going to enable that and again we have
some options coming up from those
options you expand the split evaluator
and then you can see classifier option
you expand that to again in the previous
tutorial we were interested in the
filter option but now this time we're
going for the classifier option we're
going to expand that too now here you
are seeing some parameters like debug
maximum depth number of features number
of trees seed etc remember that these
properties or attributes they're they
all belong to random forests these are
the parameters of random forests for
example I'm not going into detail but
you can choose number of features or
number of trees or even the depth of the
trees by choosing the maximum depth so
if you want to know more then you have
to take a look at random forest so these
are the parameters of random forest I'm
not going into detail so for example we
want to know we want to plot the error
rates of of this learning that we're
trying to do by changing
the no and the number of trees for
example so the default random forest
number of trees in Vega is set to ten so
we are going to change that parameter
and we will see how the error rate is
changing by plotting them in into
learning curve so we select num trees
and select it and you can see that the
default is set there ten now I'm going
to change that to several other
parameter values except ten and I'll be
back with the with the numbers when I'm
done with that
okay so I have populated the number of
trees more number of trees and the
default was set to ten now I have given
20 to 90 so this will be varied so now
we have all these number of trees for
random forests here in the generator
properties take a look at the runs that
set two from one to ten that's four
10-fold cross-validation and we're going
to use the data set called iris we go to
the way constellation package here on my
system to the program files and then we
are going to wake up Suppan six in the
data section you find I restore power so
we have done that we go to the run tab
and we press Start so the thing has
started now it's done you can see that
there were no errors we are going to
analyze tab and we click on experiment
in the comparison field we are setting
it to percentage of incorrect
predictions which is also known as the
error rate and then we do a performance
test
so now as I'm expanding this you can see
that we are having the iris dataset and
we have run random forests to predict
the labels for iris data set but we have
changed the parameter called Nam tree
or number of trees for random forests
and these are all the error rates that
are generated for 10 20 30 40 50 60 70
80 and 90 number of trees for random
forest now in order to plot the learning
curve again you need to follow the
tutorial number 36 where I have shown
how you can use these data to go to
Microsoft Excel and plot it there and
you'll find a very nice learning curve
and by that you can see where exactly
your learning algorithm is performing
the best if you're benefitted from these
tutorials on waken and other tutorials
please subscribe to the channel and
press like share as you like that will
really benefit this tutorial channel go
forward and until the next tutorial have
a nice one thank you very much
tutorial this tutorial is the
continuation of tutorial 36 where I
demonstrated how to plot learning curves
for different percentage of data size
now in this case also learning curves
can be produced for different parameters
or the change of the parameters of your
learning algorithms so I promised in
Wicca tutorial 36 that I'll be back with
these version of producing learning
curve by changing parameters of your
learning algorithm this is how you do it
so we have our way cut here and we are
going to click experimenter we are
working with experimenter and in the
experimenter we have an option
configuration mode option to set to
advanced we click on new and in the
destination it's a good practice to let
click on this particular example and
option you see that if I expand that
then we have an output file so this is
just to get output file the results
you're producing from Weka
so I'm just setting up to desktop file
called one go dark file so here all of
my results will be published so this is
a minor thing again it's it's not
necessary to understand what's going on
here it's just to save your results in
some output file so you click on ok so
that's the destination part next comes
result generator this is the important
thing to understand if you're not
following this tutorial and you can go
back to tutorial 36 the links approve
links link is provided here with the
video now in the result generator you
have the option you have some options to
choose from we're choosing the second
option cross-validation result producer
and then we are clicking left clicking
on the option here now here again you
have a bunch of options coming up number
of false output file raw code but
not interested in them were interested
in split evaluator so in this big
evaluator you can see that you have
again options click on those options and
something pops up and you see that we
have a bunch of options for the generic
object editor but we are not interested
any of them except for the classifier
from the classifier you choose let me
take it to the top you choose a meta
algorithm called filtered search you
select that and again you are left
clicking on the option here and now we
are going to set the classifier of our
of our interest which is going to be
random forests in this case and in the
end you just click OK to this window ok
to this window and ok to this window now
we have setup the result generator now
we have a generator properties here
which is disabled by default you're
going to enable that and again we have
some options coming up from those
options you expand the split evaluator
and then you can see classifier option
you expand that to again in the previous
tutorial we were interested in the
filter option but now this time we're
going for the classifier option we're
going to expand that too now here you
are seeing some parameters like debug
maximum depth number of features number
of trees seed etc remember that these
properties or attributes they're they
all belong to random forests these are
the parameters of random forests for
example I'm not going into detail but
you can choose number of features or
number of trees or even the depth of the
trees by choosing the maximum depth so
if you want to know more then you have
to take a look at random forest so these
are the parameters of random forest I'm
not going into detail so for example we
want to know we want to plot the error
rates of of this learning that we're
trying to do by changing
the no and the number of trees for
example so the default random forest
number of trees in Vega is set to ten so
we are going to change that parameter
and we will see how the error rate is
changing by plotting them in into
learning curve so we select num trees
and select it and you can see that the
default is set there ten now I'm going
to change that to several other
parameter values except ten and I'll be
back with the with the numbers when I'm
done with that
okay so I have populated the number of
trees more number of trees and the
default was set to ten now I have given
20 to 90 so this will be varied so now
we have all these number of trees for
random forests here in the generator
properties take a look at the runs that
set two from one to ten that's four
10-fold cross-validation and we're going
to use the data set called iris we go to
the way constellation package here on my
system to the program files and then we
are going to wake up Suppan six in the
data section you find I restore power so
we have done that we go to the run tab
and we press Start so the thing has
started now it's done you can see that
there were no errors we are going to
analyze tab and we click on experiment
in the comparison field we are setting
it to percentage of incorrect
predictions which is also known as the
error rate and then we do a performance
test
so now as I'm expanding this you can see
that we are having the iris dataset and
we have run random forests to predict
the labels for iris data set but we have
changed the parameter called Nam tree
or number of trees for random forests
and these are all the error rates that
are generated for 10 20 30 40 50 60 70
80 and 90 number of trees for random
forest now in order to plot the learning
curve again you need to follow the
tutorial number 36 where I have shown
how you can use these data to go to
Microsoft Excel and plot it there and
you'll find a very nice learning curve
and by that you can see where exactly
your learning algorithm is performing
the best if you're benefitted from these
tutorials on waken and other tutorials
please subscribe to the channel and
press like share as you like that will
really benefit this tutorial channel go
forward and until the next tutorial have
a nice one thank you very much
About this Video
4.65/5 Rating
Apr 29, 2026 Last updated
Related Exams
Video Description: Weka Tutorial 38: Learning Curves 2 (Model Evaluation) for Data & Analytics 2026 is part of Weka Tutorial preparation. The notes and questions for Weka Tutorial 38: Learning Curves 2 (Model Evaluation) have been prepared according to the Data & Analytics exam syllabus. Information about Weka Tutorial 38: Learning Curves 2 (Model Evaluation) covers all important topics for Data & Analytics 2026 Exam. Find important definitions, questions, notes, meanings, examples, exercises and tests below for Weka Tutorial 38: Learning Curves 2 (Model Evaluation).
Introduction of Weka Tutorial 38: Learning Curves 2 (Model Evaluation) in English is available as part of our Weka Tutorial for Data & Analytics & Weka Tutorial 38: Learning Curves 2 (Model Evaluation) in Hindi for Weka Tutorial course. Download more important topics related with notes, lectures and mock test series for Data & Analytics Exam by signing up for free.
Description
Video: Weka Tutorial 38: Learning Curves 2 (Model Evaluation) of Weka Tutorial have been curated by the Data & Analytics experts, helping you revise the topic quickly for exam preparation. Watch on EduRev.
Information about Weka Tutorial 38: Learning Curves 2 (Model Evaluation)
Here you can find the meaning of Weka Tutorial 38: Learning Curves 2 (Model Evaluation) defined & explained in the simplest way possible. Besides explaining types of Weka Tutorial 38: Learning Curves 2 (Model Evaluation) theory, EduRev gives you an ample number of questions to practice Weka Tutorial 38: Learning Curves 2 (Model Evaluation) tests, examples and also practice Data & Analytics tests.
Related Searches
Free, video lectures, pdf , shortcuts and tricks, MCQs, study material, past year papers, Extra Questions, Sample Paper, Objective type Questions, Summary, Previous Year Questions with Solutions, Weka Tutorial 38: Learning Curves 2 (Model Evaluation), Important questions, ppt, mock tests for examination, Viva Questions, Weka Tutorial 38: Learning Curves 2 (Model Evaluation), Semester Notes, Weka Tutorial 38: Learning Curves 2 (Model Evaluation), practice quizzes, Exam;