Data & Analytics Exam > Data & Analytics Videos > Weka Tutorial > Weka Tutorial 12: Cross Validation Error Rates (Model Evaluation)
Weka Tutorial 12 Cross Validation Error Rates (Model Evaluation) Video
Text Transcript from Video
hello everyone and welcome to the wake a
tutorial in this tutorial I'm going to
show you a very important thing and you
need these things whenever you are
trying to play with your data and trying
to get some information out of the data
and when you're using a careful
environment so in this tutorial I'm
going to show you how you can have the
information of every fold in a cave your
default environment for example if you
have a 10-fold cross-validation set up
then it is more likely you we will want
at some points to get the error rates of
every vault so how you're going to
achieve that so in order to achieve that
we will play with experimenter
application of waken this time rather
than playing with the Explorer tab and
we just click on that experimenter so at
first you'll see the wicca experimenter
environment is like this we have three
tabs set up run and analyze as there is
nothing here so far so the run tab is
grayed out so as everything else other
than open and new so in the setup tab we
are going to create a new experiment
environment so we created that and you
see that the grayed out is not getting
some life so for the sake of this
tutorial we'll talk about just four
panes on this particular tab called
setup the first one is experiment type
so we have told you at the very
beginning of the tutorial that we are
going to show you how you can get
information of every Fault in a careful
environment so the experiment type is
set to cross-validation but you can have
other options like train and test
percentage spit when you where you can
say that I'm going to speak my dataset
into 50 30 that means I'm sorry 70 30
that means
person of the data will be used as
training data and 30% will be used as
test data so I'm not going to do that
I'm just sticking with cross-validation
for the sake of this tutorial number of
folds you can have hundreds and
thousands of folds that depends on your
methodology and setup but for my
procedure I'm going to stick with
standard ten folds classification and
regression if your task is
classification task
you just select classification radio
button otherwise choose the regression
now the data set data set is the
particular data set you are currently
working on and on the data sets from
which you are trying to achieve those
information like error rates from every
fold in a controlled environment so
we're going to add a new data set and
I'm going to play something that I'm
currently working on with TD corpus our
fund so this data is actually an hour
file so the tarp in here is iteration
control the number of repetitions is set
to 10 as default what is this well there
are some standards of number of
reputations you need in cross validating
your training and testing stances that
means that here the 10 means that you
have 10 equal sized folds you have made
10 equal size falls out of your data set
and number of reputations means this
will be done this equal size folding
will be done up to 10 times so this is
also scumming becoming a standard to
maintain not only kid falls by
environment but also i high rated k x
environment so for the sake of this
tutorial i'm going to set a equals to 1
that means I will just run once the 10
folds of this data set called
TD corpus ARF data sets first or
algorithms first this is pretty simple
that means
if you if you chose datasets first that
means that your your data set as
continued corpus arm will be divided
into ten equal sized fold first as you
have said the number Falls to ten here
and then a classification algorithm will
be applied to it any any algorithm you
just choose that suits your purpose and
algorithms first means that the
operation will be applied on the data
set on the basis of the algorithm you
choose first and then are the wakeup
will divide that data space for you up
to ten equal sized folds so I have
anything can do anything anything answer
your purpose but my serve is for mine my
purpose is served by the IEEE data sets
first so I've chosen that algorithms
this is the classification algorithm so
you are choosing to classify training
and classify a training and test
instances or cross validated instances
from your data set so this also depends
on your task I'm going to choose nine
days for that and leaving all default
settings as it is so when they are said
you can see that now we can see the run
tab is alive we go to the run tab and we
select on start so it started it
finished and we have no errors otherwise
we should have we have if we have some
errors then we have to correct those
errors and then we have to come back to
this experiment environment now we go to
the analyze tab so see at first this is
everything here is grayed out we have
nothing to do it unless we are clicking
on experiment button so now for this set
of teacher you can do lots of things on
this app analyze pain but for this
tutorial as I've said what we are going
to do we are going to collect
information from every volt
in a careful environment okay so we're
going to show our results in this way
under rows we are going to put the
number of folds so that's going to be 1
2 3 up to 10 and on the column we're
just going to have one information that
percentage of error or in other words
that's also called error rates so on the
road button a row called row a set up we
we are just setting it up to fold
because we want a rose as fold we select
that and on column we're selecting our
percentage in current that's the error
rate so you can see that you have number
of options here number of options
anything you can you want to get the
information from a particular voltage
you can select that so we select that
and that after that to selecting that
you can see the available result says
one to ten so these are the ten folds of
your data sets and these values 2480 131
11 up to 2007 these are the error rates
of every volt so now with these error
rates you can see if your classification
algorithm is overfitting or not what's
that based classifier you have designed
and so many things you can do with this
percentage of error or error rates or
from every fold from a data set so you
can go back to all school books of
machine learning and see what you can do
what are you actually what you can
actually do with the error rates from
every fold
tutorial in this tutorial I'm going to
show you a very important thing and you
need these things whenever you are
trying to play with your data and trying
to get some information out of the data
and when you're using a careful
environment so in this tutorial I'm
going to show you how you can have the
information of every fold in a cave your
default environment for example if you
have a 10-fold cross-validation set up
then it is more likely you we will want
at some points to get the error rates of
every vault so how you're going to
achieve that so in order to achieve that
we will play with experimenter
application of waken this time rather
than playing with the Explorer tab and
we just click on that experimenter so at
first you'll see the wicca experimenter
environment is like this we have three
tabs set up run and analyze as there is
nothing here so far so the run tab is
grayed out so as everything else other
than open and new so in the setup tab we
are going to create a new experiment
environment so we created that and you
see that the grayed out is not getting
some life so for the sake of this
tutorial we'll talk about just four
panes on this particular tab called
setup the first one is experiment type
so we have told you at the very
beginning of the tutorial that we are
going to show you how you can get
information of every Fault in a careful
environment so the experiment type is
set to cross-validation but you can have
other options like train and test
percentage spit when you where you can
say that I'm going to speak my dataset
into 50 30 that means I'm sorry 70 30
that means
person of the data will be used as
training data and 30% will be used as
test data so I'm not going to do that
I'm just sticking with cross-validation
for the sake of this tutorial number of
folds you can have hundreds and
thousands of folds that depends on your
methodology and setup but for my
procedure I'm going to stick with
standard ten folds classification and
regression if your task is
classification task
you just select classification radio
button otherwise choose the regression
now the data set data set is the
particular data set you are currently
working on and on the data sets from
which you are trying to achieve those
information like error rates from every
fold in a controlled environment so
we're going to add a new data set and
I'm going to play something that I'm
currently working on with TD corpus our
fund so this data is actually an hour
file so the tarp in here is iteration
control the number of repetitions is set
to 10 as default what is this well there
are some standards of number of
reputations you need in cross validating
your training and testing stances that
means that here the 10 means that you
have 10 equal sized folds you have made
10 equal size falls out of your data set
and number of reputations means this
will be done this equal size folding
will be done up to 10 times so this is
also scumming becoming a standard to
maintain not only kid falls by
environment but also i high rated k x
environment so for the sake of this
tutorial i'm going to set a equals to 1
that means I will just run once the 10
folds of this data set called
TD corpus ARF data sets first or
algorithms first this is pretty simple
that means
if you if you chose datasets first that
means that your your data set as
continued corpus arm will be divided
into ten equal sized fold first as you
have said the number Falls to ten here
and then a classification algorithm will
be applied to it any any algorithm you
just choose that suits your purpose and
algorithms first means that the
operation will be applied on the data
set on the basis of the algorithm you
choose first and then are the wakeup
will divide that data space for you up
to ten equal sized folds so I have
anything can do anything anything answer
your purpose but my serve is for mine my
purpose is served by the IEEE data sets
first so I've chosen that algorithms
this is the classification algorithm so
you are choosing to classify training
and classify a training and test
instances or cross validated instances
from your data set so this also depends
on your task I'm going to choose nine
days for that and leaving all default
settings as it is so when they are said
you can see that now we can see the run
tab is alive we go to the run tab and we
select on start so it started it
finished and we have no errors otherwise
we should have we have if we have some
errors then we have to correct those
errors and then we have to come back to
this experiment environment now we go to
the analyze tab so see at first this is
everything here is grayed out we have
nothing to do it unless we are clicking
on experiment button so now for this set
of teacher you can do lots of things on
this app analyze pain but for this
tutorial as I've said what we are going
to do we are going to collect
information from every volt
in a careful environment okay so we're
going to show our results in this way
under rows we are going to put the
number of folds so that's going to be 1
2 3 up to 10 and on the column we're
just going to have one information that
percentage of error or in other words
that's also called error rates so on the
road button a row called row a set up we
we are just setting it up to fold
because we want a rose as fold we select
that and on column we're selecting our
percentage in current that's the error
rate so you can see that you have number
of options here number of options
anything you can you want to get the
information from a particular voltage
you can select that so we select that
and that after that to selecting that
you can see the available result says
one to ten so these are the ten folds of
your data sets and these values 2480 131
11 up to 2007 these are the error rates
of every volt so now with these error
rates you can see if your classification
algorithm is overfitting or not what's
that based classifier you have designed
and so many things you can do with this
percentage of error or error rates or
from every fold from a data set so you
can go back to all school books of
machine learning and see what you can do
what are you actually what you can
actually do with the error rates from
every fold
About this Video
4.89/5 Rating
Apr 29, 2026 Last updated
Related Exams
Video Description: Weka Tutorial 12: Cross Validation Error Rates (Model Evaluation) for Data & Analytics 2026 is part of Weka Tutorial preparation. The notes and questions for Weka Tutorial 12: Cross Validation Error Rates (Model Evaluation) have been prepared according to the Data & Analytics exam syllabus. Information about Weka Tutorial 12: Cross Validation Error Rates (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 12: Cross Validation Error Rates (Model Evaluation).
Introduction of Weka Tutorial 12: Cross Validation Error Rates (Model Evaluation) in English is available as part of our Weka Tutorial for Data & Analytics & Weka Tutorial 12: Cross Validation Error Rates (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 12: Cross Validation Error Rates (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 12: Cross Validation Error Rates (Model Evaluation)
Here you can find the meaning of Weka Tutorial 12: Cross Validation Error Rates (Model Evaluation) defined & explained in the simplest way possible. Besides explaining types of Weka Tutorial 12: Cross Validation Error Rates (Model Evaluation) theory, EduRev gives you an ample number of questions to practice Weka Tutorial 12: Cross Validation Error Rates (Model Evaluation) tests, examples and also practice Data & Analytics tests.
Related Searches
Viva Questions, shortcuts and tricks, Weka Tutorial 12: Cross Validation Error Rates (Model Evaluation), Semester Notes, Important questions, pdf , Weka Tutorial 12: Cross Validation Error Rates (Model Evaluation), practice quizzes, MCQs, video lectures, ppt, Weka Tutorial 12: Cross Validation Error Rates (Model Evaluation), study material, past year papers, Extra Questions, mock tests for examination, Summary, Sample Paper, Exam, Free, Previous Year Questions with Solutions, Objective type Questions;