Taught by Patrick Hebron at ITP, Fall 2016
We will start today's class by answering any questions that arose from your final project work or continued explorations of TensorFlow, LaunchBot and Docker.
What is Learning?:
In Week 1, we started our study of machine learning with a discussion of the questions listed below:
We use the terms learning and intelligence often, but what do they mean?
- When a squirrel buries an acorn, does that require intelligence or mere instinct?
- What about when an ape uses a stick to extract ants from an anthill? Does that require intelligence?
- Is intelligence required for a calculator to perform a complex mathematical operation?
- What's the difference between rote operations, instinctual behaviors and learned behaviors?
- Does learning begin at birth or can it be inherited?
- When we say intelligence do we really mean human intelligence?
- Is it possible to provide universal definitions for the concepts of learning and intelligence?
With the semester now coming to a close, how has our study of machine learning impacted our understanding of the nature of learning and intelligence in general?
From Iteration to Reflection on Process:
Project Development Exercise:
- What did you learn in the process that might have changed your initial approach?
- What would you do exactly the same way?
- How were your high-level concepts impacted by low-level technical or design considerations?
- Think about the tools you used to develop your project:
- In what ways did these tools aid your creative process?
- In what ways did they hinder your process?
- How did these tools shape your overall vision for the project?
- Were there things about these tools that seemed more challenging, confusing or limiting than was necessary?
- Were there things you were hoping to achieve that you couldn't because of limits in the underlying tools?
- What would have been required to get around these limits?
- How did implementing machine learning algorithms from scratch inform your project ideas and process?
- How did working with higher level machine learning tools inform your project ideas and process?
- When is a project finished? Is this project finished?
- Prepare you project for final presentation. Be ready to:
- Discuss your project and development process
- Give a brief demonstration of the project
- Exchange feedback with your classmates