About

About Iris Machine Learning Demo is a web application for demonstrating machine learning task on the web using multiple classification model.

I developed this as a tool for demonstrating the possibility of formulating and deploying multiple machine learning model on Responsive Web using simple JavaScript and rudimentary Web skill to learners and students..

Iris Dataset

The demo uses the well known classic Iris Dataset by R.A Fisher and popularised by Duda and Hart. The main reason Iris dataset is chosen because of its popularity being used in beginners' machine learning course and it can clearly demonstrate multivariate classification without additional complexities.

The dataset can be downloaded from:

  1. UCI Machine Learning Repository - Iris Dataset
  2. Kaggle - Iris Flower Dataset

Classification Algorithm used

Currently the web application used the following algorithm

  1. Multilayer Perceptron Neural Network
  2. J48
  3. PARTS
  4. RandomTree

The Multilayer Perceptron Neural Network model was trained using Keras, and made available on the web with the help of TensorFlow.js library.

The J48, PARTS and RandomTree are trained with WEKA 3 : Machine Learning Software in Java and deployed/transfered to the web using basic Javascript skills.

Refer to this YouTube video (Coding Malaya channel - Malay Language) on the basics of transfering WEKA output to JavaScript using (very) basic programming skills.

Author

Mohammad Hafiz bin Ismail (mypapit@gmail.com)

For any enquiries, please send an email to deeplearn@uitm.edu.my

How to cite this web application

If you are referencing this web application in your work, please cite it as :

Ismail, M. H. (2022). Iris Dataset Machine Learning Demo - a Web Application Demonstrating Multiple Machine Learning Algorithm for Classification task. Retrieved , , from: https://demo.mobilepit.com/ai/iris/.

You can also import the citation from this Bibtex file.