Debrief of Week 1

class: center, middle, inverse, title-slide

# Debrief of Week 1
## AI / Machine Learning<br>Spring 2021
### Ken Arnold
### 2021-02-15

---

## Logistics

* [attendance sheet](https://docs.google.com/spreadsheets/d/1FCogt4a5yv3_jauwA6EUkzCYrVV0Yww6OoLplM7Tq54/edit#gid=112619962): <http://tiny.cc/cs344here> or link on Moodle.
* Reading quiz *today*
* Ethics readings for *Wednesday*
* Discussion forum
* In classroom all week

---

## Scripture on Intelligence

.can-edit.key-scripture-intelligence[
* God *expects* us to act intelligently, mirroring how he himself does.
  * example: [**Psalm 103** NIV](https://www.biblegateway.com/passage/?search=psalm+103&version=NIV)
* Psalm 139
  * The intelligence of even a human baby is pretty impressive
  * God gave us purpose
* Genesis 1
  * Our intelligence and creativity is an image of God's
]

---

## Reflection

.absolute.right-1.top-1[Next one due Wednesday!]

What could you have referenced for each of the topics?

.can-edit.key-reflection-refs[
* Tech
  * ...
* Community
  * ...
* Context / History
  * ...
]

???

* Tech
  * Quizzes (Py, reading)
  * Lab 1
  * You can point to evidence that you understand Github, e.g., a link where the changes you pushed can be seen
* Comm
  * Teachable Machine first-day activity

---

## Reflections: Tech

Which communicates more about what you learned?
<br><br>
.pull-left[
"I had some trouble with my lab partner getting things to work. Especially trying to figure out the number of test and valid cases. We eventually figured it out"
]

.pull-right[
"didn't learn much implementation, but I made hypotheses about the neural nets we created in lab and analyzed them afterwards to verify that: a. Using more layers reduces the runtime (though actually only two layers are necessary) and b. reducing the amount of training data leads to a higher error percentage (such as by increasing the validation % or changing to a more unique/diverse classifier). This went well; I found the lab easy to understand after watching the video."
]

.can-edit[
]

---

## Reflection: Tech

> I didn't understand which error rate to use since it gives two of them. But I asked about it and found out that we should use the last one and we'll learn why in a few weeks.

Who else was wondering this?

---

## Reflections: Justification

> I think that I deserve a 95/100 because of the time/effort I took.

Is this a good justification?

.can-edit[
* ...
]

???

* What was good?
* What was lacking?

---

## Reflections: Community

> I talked with a friend about course work. It went well

* What does this tell us? What more could it say?
* What other ways can you help make our class a learning community?

.can-edit[
* Piazza (ask, answer)
* Discussion Forums
* Collaboration in lab / hw
* Speaking in class
]

---

## Peers

Your peers provide:

* **Collaborators**. You'll be able to do great projects together.
* **Learning community**. I may be one of your last official teachers. But you most keep learning.
* **Discussion partners**. Your peers bring valuable perspectives.
* **Practice analyzing other people's work**. It will be most of what you do. So we must practice it. We will do some peer feedback during our time together.

---

## "I don't know"

* Say this often.
* I will too.
* Ask questions on Piazza.
* When you've figured it out, communicate (e.g., a blog)

---

## Reflections: Context / History

> In our capstone class, I brought up some ideas about how AI relates to technology. We were talking about the limits of software and I proposed some ideas about how AI could be limited.

---

## Resources

<https://cs.calvin.edu/courses/cs/344/ka37/#resources>

---

## Homework 1

.can-edit[
Build and evaluate a classifier to determine if a photo is taken of the inside vs outside of a restaurant. Don't scrape photos; use the Yelp Academic Dataset.

### Steps:
* ...
]

---

## Lab 1

See [this example](https://colab.research.google.com/drive/1kf1dwlCbqoL01RbVa7kPa1rjuTi1hyuA?usp=sharing) (<https://tiny.cc/jxhltz>)

---

## Running Code

.pull-left[
* lab machines
* Google Colab
* your own computer
* Another cloud provider?
]

.pull-right[
* Teams screen-share
* Github commit frequently
* ...even if you're using Colab.
]

Colab preamble:

```
try: import fastbook
except ImportError: import subprocess; subprocess.run(['pip','install','-Uq','fastbook'])
```

---

## What is this course?

This is a course on developing AI systems using machine learning, with a particular emphasis on deep neural networks.

### Objectives

By the end of this course, you will demonstrate growth in your ability to:

- *Design, implement,* and *analyze* some state-of-the-art types of data-fueled intelligent systems
- *Design* and *analyze* how these intelligent systems relate with people, individually and collectively, through a Reformed Christian lens.
- Communicate the above with peers and members of the general public
- Contribute to a team that learns: identify what you need to learn, reflect on your own learning, and help others learn

---

## Real objective

.center[
**You graduate and do amazing work in AI.**
]

So, you should be:

* self-directed, collaborative learners
* able to be immediately productive at some kinds of ML problems
* in the habit of practicing the integrity and perseverance needed to work in a team

---

## Why not just learn this on your own?

This course gives you:

* Structure
* An instructor
* Peers
* Freedom to ask stupid questions