Wednesday, March 20, 2013

Tutorial on creating a confusion matrix using Orange

Confusion Matrix

The following video shows a tutorial on how to create a confusion matrix using Orange. I used the same dataset as used on an existing confusion matrix example on Youtube, as this is only one available on Orange which effectively generates a confusion matrix. The screen recording software is not the best as I had to record the video and audio separately and this may have caused some lag. I tried at least 3-4 different kinds of recording tools but had compatibility issues with Windows 8.

This is a very useful tool and can be extensively used in analytics. Hopefully the video is clear enough or I can write about this explaining it step by step. I did the tutorial to visualize the results and share with the class on how to use this feature in Orange as I learnt this from various tutorials online.
It uses a dataset which categorizes 3 different types of flowers- iris setosa, iris versicolr and iris virginica in terms of sepal length and width and petal length and width. The confusion matrix shows how sometimes the system might wrongly categorize these flowers based on closely matching attributes even with a well categorized excel file.

No comments:

Post a Comment