Communications

ESE 2000 – This is all, see you soon (8/25)

Dear all,

I should apologize for the spam. You should have received 4 emails from me not counting this one:

  • (2) Information about the course and Instructor. In particular, I asked you to read the logistics page and acquaint yourself with expectations and policies. Also, make sure you check the course’s calendar.
  • (3) A specific invitation to check the grading policy. 
  • (4) A study guide for Weeks 1 and 2. This is very important. There is a lot going on.

This is all for now. I will see you on Wednesday. On that note, please find attached the slides I will use in the first two lectures.

Best,

Alejandro

ESE 2000 – Week 1 Study Guide (8/25)

Dear all,

As we gear up for Weeks 1 and 2, I wanted to help you out with a study guide. During this week and the next we will be working on Labs 1A and 1B. The purpose of Lab 1A is to introduce the components of an AI system and the purpose of Lab 1B is to study gradient descent and stochastic gradient descent.

We have only one lecture this week and one other lecture next week.  This is good because it helps us to start slowly. This is what you should do during these first two weeks:

  • (3) The reports for Lab 1A is due on Tuesday September 2 and the Lab 1B report is due on Monday, September 8.  These labs are intended to allow you to get up to speed. They contain simple tasks and the reports are minimalistic. Each of them can be completed in 3-4 hours. That said, I know that they will be confusing and take much more than that for some of you. This is why we are starting slowly! 

As you attempt to work on Labs 1A and 1B I can foresee that some of you may have difficulty with matrix algebra and Pytorch. If you do, we have materials available to help you out:

  • (4) If you have trouble with matrix algebra, read Appendixes A and B in the Lab 1A assignment. You just need to either learn or remember a few simple definitions of what it means to multiply vectors with matrices.
  • (5) If you are having trouble with Pytorch, read Lab 1X. This is a tutorial that is designed to teach you the basics of how to get started with Pytorch.

Finally, a few of you will have trouble with software installation. The instructions are very simple but there is always potential for a hitch. If that is the case:

  • (6) Go to office hours (they will be posted soon). Better yet, save your time and talk with one of your classmates. We are happy to help but this really is very simple. If it is not working for you it is because of a silly mistake.

Best regards,

Alejandro 

Welcome to ESE 2000 – Grading Policy (8/25)

Dear all,

Please make sure to check the grading policy in the logistics page of the course’s website. I will NOT discuss this in class and future claims of ignorance will fall on deaf ears.

Take notice of the mandatory attendance policy and the midterm dates. I am giving a very early warning on these dates. Please plan your travels accordingly.

Best,

Alejandro

Welcome to ESE 2000 – Course and Instructor Info (8/25)

Dear all,
Four years ago, I convinced myself that it was possible to teach the basic mechanics of AI to first year students. I just needed to walk away from some of my preconceptions about statistics and learning theory and focus on linear algebra and optimization. The result of that was ESE 2000 which, over the years, has grown to cover quite a lot of ground. Within 12 weeks you will learn about the three architectures that underlie almost all of the AI systems in use these days (CNNs, GNNs, and Transformers) and about the four ways in which almost all of the AI systems in use are trained (stochastic gradient descent, recursive generative sequences, generative diffusion models, and reinforcement learning). There is, of course, a lot more to AI than these basic introductions but at least you will end up knowing more about AI than Sam Altman. Which is not a very high bar to clear, but not a meaningless one, either. 
ESE 2000 is run through this public website. Labs and solutions are posted here along with logistics information. I need you to read the logistics page before our first meeting.  There is information in this page about my expectations, due dates, grading, etc, that I am NOT, repeat NOT, going to discuss in class. I’d rather we have an interesting day than bore you with housekeeping information.

In particular, please check the calendar so that you can organize your study schedule and be aware of midterm dates. Also, because this is unusual, read the policy forbidding the use of electronic devices during lectures.

If you want to know about me, please check my group’s website. We are known for our work on the foundations of convolutional information processing in general and graph neural networks in particular. We have received awards for our work on stability and scalability. We have also been pioneers in the introduction of requirements in AI. On the application side, we have written seminal papers that develop the use of AI in multiagent robot systems and wireless communication networks. 

I take pride in being a dedicated teacher and mentor. Although I have not received any teaching award in 10 years, this is mostly explained by the fact that it took me 8 years after I started teaching to collect the two awards that Penn offers. My former doctoral students and postdocs are an accomplished bunch. About half of them are professors, the other half work in the usual tech companies, and two are successful entrepreneurs.

I am also husband to Gabriela and father to Miranda (Penn class of 2020), Guille (Penn class of 2022) and Ariel (Penn class of 2031, I hope) and contrary to what it may look like at times, I do not take myself too seriously.

Best,
Alejandro

Welcome to ESE 2000 (8/25)

Dear all,

My name is Alejandro Ribeiro and I am in charge of teaching ESE 2000 this coming Fall. I am very much looking forward to working with you. I will follow up later today with a couple of more informational emails.

In the meantime, please visit the course’s webpage (https://ailab.seas.upenn.edu) and navigate to the labs page. Read about software requirements and install the necessary components if you have to. If this is a challenge, we will soon have office hours to help you out. 

I further recommend that you read the assignments for Lab 1A and Lab 1B. We will work on Lab 1A this week and on Lab 1B next week. 

Best regards,

Alejandro