Examples of Classroom Practices

This document shows some examples of embodied activities from our courses.

Embodied Activity: RNN¶

source

RNN is like your brain reading a sentence word by word.

• Input at each time step: The current word you read

• Hidden state: Your current mental understanding

• Output: Your interpretation, reaction, or prediction at that point in time

Two rows of students:

• Front row = input layer (observations at each time step)

• Back row = hidden state at each time step

• Each column is a time step (0 through 4)

• So we’ll have 4 front-row students: $x_0$ to $x_3$

• And 4 back-row students: $h_0$ to $h_3$

1. At time step 0:

   • Front-row student $x_0$ gets a word

   • They pass it to the back-row student behind them ($h_0$).

   • $h_0$ is initialized with an image if you want to make it interesting

2. At time step 1 (and beyond):

   • The front-row student (e.g., $x_1$) gets a new word

   • The back-row student (e.g., $h_1$) receives:

     • The current input from the front-row student (e.g., $x_1$)

     • Whatever “memory” is passed from the previous hidden state (e.g., $h_0$)

     • $h_1$ combines this (e.g., by writing a summary phrase or combining keywords).

3. Repeat until time step 3 or 4.

4. Final time step: $h_3$ summarizes what they remember (e.g., predicts next word, gives the “mood” of the sentence, etc.)

Embodied Activity: Markov Models of Language¶

Source

Each of you will receive a sticky note with a word on it. Here’s what you’ll do:

• Carefully remove the sticky note to see the word. This word is for your eyes only; don’t show it to your neighbours!

• Think quickly: what word would logically follow the word on the sticky note? Write this next word on a new sticky note. You have about 20 seconds for this step, so trust your instincts!

• Pass your predicted word to the person next to you. Do not pass the word you received from your neighbour forward. Keep the chain going!

• Stop after the last person in your row/table has finished.

• Each row/table reads their generated sentence.

Whichever row generates the most captivating sentence will get XXX.

Here are some example sentences generated by students.

• Row 1: We trust God is student study Hard Rock Johnson buys candy

• Row 2: Is student study hard rock johnson buys candy

• Row 3: Data center tendency increase fluctuation during afternoon tea party invitation acceptance

• Row 4: The size of candy sweet pudding gross grossy yard fence (??????) 😀

• Row 5: We play games are mingle cuddle pet cat licking ice cream with you eat.

• Row 6: What are you talking about?

• Row 7: I like to eat ramen spicy story time series in the game time off

Course Structure: Spiral Learning¶

The spiral learning approach is described in this paper, below is the abstract reproduced:

Introductory courses often present techniques in a linear sequence, resulting in a steep learning curve that can overwhelm students and limit the time for experiential learning through course projects. To address this, I restructured the course using a spiral approach, presenting concepts in three iterations. Each iteration delves deeper into the material and introduces complex computational topics progressively. This method includes a built-in repetition mechanism that reinforces learning and enhances understanding. Moreover, this approach allows time for hands-on projects that apply theory to real-world scenarios, helping students better understand the course materials. The spiral approach was implemented in an ML course at a local university, resulting in positive student feedback and improved course retention rates.

Traditional linear approach

Spiral learning approach