You are on page 1of 5

Artificial intelligence lab

Project summarization

Submitted by: Submitted to:


Sikder Md Mahbub Alam Jargis Ahmed

Id : 183002005 lecturer
Section : Da CSE, GUB
Dept : CSE
abstract:

● To understand WEKA tool usage to analysis training data.


● To understand how to use Java programming to perform data
analysis in WEKA.

introduction :
i use image segmentation file to do this project. Here in WEKA i perform
analysis on training data- "Decision Tree" usage using GUI. i use j48 trees
and multilayer perception algorithm . changing the fold number , accuracy
and hidden layer to find out different result and ultimately drawing some
graph to see how it’s working against changing those value .

image segmentation:

● Segmentation is an important stage of the image recognition system,


● because it extracts the objects of our interest, for further processing
such as description or recognition. ... Segmentation techniques are
used to isolate the desired object from the image in order to perform
analysis of the object.
● The instances were drawn randomly from a database of 7 outdoor
images.
● Classes : brickface, sky, foliage, cement, window, path, grass.
● In my file there are 19 attributes .
decision tree:
using j48 algorithm : changing the fold number against accuracy

so we can see from fold number 2 to 4 it’s increasing and then with the
increasing of fold number it’s decreasing again and again from fold number 6
it’s increasing again , so we can see a gradual increment and then a gradual
decreament .
Neural network : changing hidden layer against accuracy (all the time fold
number 10)

so we can see with the increasing of hidden layer the accuracy is increasing

Neural network : changing training time against accuracy (all the time fold
number 10) c

so we can see with the change of training time accuracy is increasing but
after a certain point at 200 the accuracy is all the same
ccura

You might also like