 So, after you discussed your expectations, let me also share my expectations. I expect all of you at the end of this course to be able to constructively use deep learning to solve real world problems. You will need more courses than just this course, if you want to be a deep learning researcher. It's an introduction to deep learning course, but you should be at the level where you can have fun with it and use it and really solve problems with it. I also hope that you'll be able to see through the hype. Deep learning is not the solution to the pram of how intelligent or the universe works. So I hope you'll see through the hype. And lastly, my expectation is that you will help us to make this course better. Now, as a short edition here, the pandemic is difficult for all of us. It's very difficult for me. I am certain it's going to be difficult for a lot of you. I want to emphasize that the University of Pennsylvania has a lot of useful resources there. They will all be listed in the collabs. If you're at UPenn, use them. They're there for you. But now, let's directly talk about the curriculum. It's an interesting way how we structured that, because in week one, today, we will do alpha zero. In a way, what we try to do is do an advanced topic that uses reinforcement learning alpha zero that gets you to see the big picture of things. Because it's much more fun working on the details and trying to understand the minutia of the system once we have this intuition of how it all works together. So in that sense, that first week serves as an index into the rest of the course where we see the basic logic, we see the big structure. And that will make us be more interested in all the components that are there. Now, in the next four weeks, weeks two to four, we will be talking about the nuts and bolts of deep learning. How do we set up neural networks? How do we use PyTorch? How do we deal with hierarchical, interesting systems? How do we optimize things? How do we make sure that it's properly regularized? Those are the nuts and bolts. That is what makes deep learning work. A lot of it isn't as fancy as say alpha zero, but a true mastery of the nuts and bolts is what makes you effective at solving problems. We will also hopefully get you to understand how to debug the problems that will invariably happen when you use deep learning and that will arguably, in your career, take much more time than the actual building of deep learning systems. Then in weeks six to eight, we will talk about computer vision. We'll learn about confnets, transfer learnings, learning generators of images. Then in weeks nine and 10, we will talk about natural language processing. We'll talk about text, about recurrent neural networks, and how to put it together to work with interesting text data. Then in weeks 11 to 12, we'll talk about reinforcement learning, about deep learning systems that interact with games or more interestingly, with the real world. And then lastly, in weeks 13, we will recapitulate what we have learned so far. So there, we will basically put all the things that we learned so far together to make sure that the things that we learned, that we really don't forget them and that we have full mastery of them. And then for the rest of the term, we'll be working on projects where guided by your TAs, you will all choose some really cool deep learning problems and solve them in the time that you have. So some of you might be interested in, well, I'm actually interested in say reinforcement learning, can I kind of tune out of the early parts of the course? Not like it's reinforcement learning that I'm interested in. The short answer is no to that. And in more detail, deep learning consists of lots of tricks. As we will see today, for example, we use convolutions that are taught as part of computer vision for reinforcement learning tasks. So all these ideas that we have anywhere in deep learning reoccur just about anywhere. And therefore, we want to make sure that as many of these principles of these tricks are taught that you can use them. And they will occur everywhere, both before the lectures that might be most interesting to your topic of interest. And they will reoccur earlier or later in the course. It's really important that you work in great detail through all of them. Without the earlier lessons, you will also miss crucial details in the later ones. So back to you. Discuss with your part what you're most hoping to get out of this course. Which week do you think will be your favorite? And make sure you share that with us. Come back in 10 minutes.