IRIS Flower Classification Prediction Application

Coursera: Developing Data Products Student
July 15 2014

Introduction

Welcome to the IRIS Flower Prediction Application. The purpose of this application is to help any flower collector or hobbyist to accurately classify their IRIS flower collection.

Features of this application:

Application has been trained on a wide variety of IRIS flower types
Application uses a modern machine learning algorithm namely Random Forest
Application has an almost perfect prediction record as measured by the confusion matrix
Application is very easy to use

Instruction

It's so easy to use this application. These are the five steps to follow:

#1 Take the following measurements of your IRIS Flower: Sepal Length and Width, Pedal Length and Width
#2 Enter those 4 values into the left panel either directly or using the provided controls
#3 Press the Submit button
#4 Ensure that your data entered is accurate in the main panel and view the results of the Classification Prediction
#5 Go back to #1 if you have another IRIS flower you wish to classify.

Random Forest Trained IRIS Model

The plot shows the Error Margin which approaches zero.

plot of chunk unnamed-chunk-1

Random Forest Confusion Matrix

This table shows the confusion matrix between different IRIS Species and the classification error.

                Iris-setosa Iris-versicolor Iris-virginica class.error
Iris-setosa              50               0              0        0.00
Iris-versicolor           0              47              3        0.06
Iris-virginica            0               3             47        0.06