Weka Tutorial 20: Attribute Selection with Knowledge Flow Environment (Data Dimensionality) Video Lecture | Weka Tutorial - Data & Analytics

hello YouTube in this tutorial I'm going
to show you how you can use the Waker
knowledge flow environment application
to see the selected attributes by rapper
method so if you do not know how repre
method of attribute selection works then
you can go back to some of my tutorials
where i explained how you can do that
with wake I Explorer but this is a user
request where a user has requested to
show this exact same thing with waken
knowledgeful environment so I'm going to
start with waking knowledge flow
environment so as I open the knowledge
for environment at first I have to take
the data source and I'll put the our
floater on the layout second I have to I
have to put the class label assignor so
the class level assignor is on the
evaluation tab class assignor is
actually telling you which which of that
should be obscene your attribute list of
the data set is going to be the class
attribute so we place that on the layout
and then we will go to the filters tab
and you'll see that we will select the
attribute selection and put that on the
layout now visualization we need to see
what attributes have been selected so we
will go for a text viewer and put that
on onto the layout so this is our
primary setup now we have to put a link
between those selected items on the
layout we click on the our floater we
select data set connect it with the
class asana will you right-click on the
class assignor
you select the data set and put it to
attribute selection we right click on
that group selection select the data set
and fit that to text viewer so that's
fairly easy now we have to configure the
class designer so I'm going to right
click on the class designer you can see
this is a classic sonic customizer you
go down you can see the class column is
set to last
is the last attribute of my attribute
list in the data set is going to be the
class attribute so at this moment I'm
clicking ok but I'll show them later why
I put that to lust now in the attribute
selection as I told this is going to be
a rapper method we are double clicking
on the attribute selection icon here and
we choose evaluator as the classifier
subset eval this first is the default
search and I'm happy with that
now I'm going to click on the classifier
subset idle option and you know that my
favorite classifier so far is because of
its simplicity is the night pace so I
click on the base classifiers and select
the night bass holdout file is going to
be the default ones so I'm not going to
change that I click OK here and I click
OK for filter options now I'm going back
to the file I'm going to test with this
is the file I'm going to load in my our
floater
it has 39 attributes and you can see
that the last attribute is the class
attribute so that's why in the week and
knowledgeable environment on the class
assignor I have chosen the last as plus
column so I click OK
so I am going to select that our file I
right click on our floater
select configure I go to the place where
I'm keeping my file so this is the file
I was talking about and it has been
loaded so now I right-click and I click
on start loading you can always go back
to John you can see that everything has
wore fine so our float a floater is
running for seven seconds attribute
selection is running for 10 seconds now
so it's okay we go to the top now
everything is done for you directly my
text viewer we click on show results and
you can see that out of 39 attributes 9
is performed with this subset of
attributes the best so we have six
attribute selected by nine based
classifier with one class attribute so
this is the attribute
list that has been provided to you by
nine days so in this way you can also
apply the filter method to find out the
attributes that have been selected from
a list or pool of attributes I hope that
helps thank you very much for watching