ABOUT THE INSTRUCTOR
- Machine Learning Research Scholar, IISc Bangalore
- Over 2 years of work experience as a Data Scientist
ABOUT THE COURSE
This is not a course but an interaction. This interaction is about learning new and creative things. Through this medium we will explore creative ways of making a machine ‘learn’ through data and make the machines interact with us in more meaningful ways than before. It’s about making the machine understand the difference between the images of a rocket and a ball pen, the difference between a sad and a happy face, and what do we mean when we say up, down, right or left, to a machine?
Moreover, it’s also about learning why it is really hard for a machine to understand such instructions when it is pretty easy for a three year old child. Why does the machine sometimes act foolish about such things when it is much easier for us?
Overall, we will explore how we can use ‘machine learning’ (ML) to make a web based interactive tool on our own with an integrated intelligence in it.
All of this can only be done through a set of activities rather that a set of boring lectures. Hence, we plan to do the following curriculum.
Learning about Machine Learning and platforms for ML interactions
- Understand what is AI/ML how we can use them on various creative learning platforms
- Learn about terms and basic definitions in the ML community
- Introduction to creative learning platforms
- Learning about Canvas, shapes, and many inbuilt functions of the platform
- Learning about color codes, primitive shapes and their movement
14th June (2-3:30 PM IST)
Create a framework and a design for implementing ML model
- Create a basic design on the Canvas using our knowledge about shapes from Activity 1
- Understand the importance of the relative positioning of such shapes on the canvas
- Learn how to extract mouse cursor data (x-y coordinates) to bring dynamic changes in such shapes on the Canvas
- Learn how to use input data from the keyboard to create interactions in the Canvas
Learning about various Machine learning models and coding practices to create such models
- Learn about Image recognition, video web cam classifier, sound classifiers
- Understand the coding practices needed to build such models
- Build our own models through live coding
- Test the models by providing data to them
- Learn about pros and cons of different machine learning models
Integrating our ML models into our designed frame work
- Integrate ML models from Activity 3 into out designed framework from Activity 2
- Understand common nuances such as errors and predictions to create more such Machine Learning frameworks in the future
- Host our work on Github for sharing it to people