# 🫂 Advice

We asked current tutors to share study tips based on how they studied for DSC 40A when they took it. Here are their raw, unedited answers.

### Yosen Lin's Advice

- Form study groups when studying for exams and when doing homework assignments. There are times when your friends have a different approach from you. You can learn from them if their approach is more efficient. If it turns out that your approach is more efficient, then you earn an opportunity to teach your friends. Not only will explaining your approach help your friends learn, it also reinforces and solidifies your understanding of the concepts applied to the problems that you are trying to solve.
- When stuck with a specific problem, think about the simpler case of that problem:
- Ex: CAVOCADO problem. If you are having trouble proving the given answer, start from computing the number of permutations of CAVO, then CAVOC, and so on.

- Do practice problems and show up at office hours!

### Utkarsh Lohia's Advice

- Go over the lecture slides/ your notes to review all the concepts. If there’s any concept you have trouble with, rewatch that lecture to get a clearer idea.
- If you still have trouble grasping the concept, go to office hours.

- Look over and redo all homework and groupwork problems in order of their release.
- If your answers and the solutions do not match, ask on Ed / Office hours whether your approach is correct

- Do practice exams in an exam-like environment. No music, a set time limit, no external distractions(phone, iPad), and do the exam on paper.
- If your solutions do not match up with the actual solutions, try to go over your explanations and see where you went wrong.

- If you are not confident in any of the topics, redo problems from those topics from the practice site.

### Zoe Ludena's Advice

I know generically a tip for doing better on exams is “looking at lecture and doing practice problems,” but I add a little bit of spice to these ideas. Looking over the lecture is fine, but trying to re-group topics together on your own and writing it out can really benefit you! For example, I would take the time to group together my notes on empirical risk and loss. Perhaps add in some notes if I struggled with derivative rules. Another way I study is working with my friends. We would go over practice problems, class problems, and lectures together. Sometimes I struggle with things that others do not or am better at those things than them. Teaching people is a fantastic way to demonstrate your understanding of a topic. Teaching anyone can help, even people who do not know the material! It forces you to know the topic.

### Mert Ozer's Advice

First, gather all the resources and practice problems at your disposal. Then, start studying with the topics that intimidate you the most or the question formats that challenge you the most. If you manage to review all your resources before the exam, fantastic. Even if you don’t, keep in mind that your fearless preparation has already put you in a strong position. 🚀

### Harshi Saha's Advice

- Go through items in order of when they were released / due:
- Read through annotated lecture slides for each week
- Then read through the solutions on that week’s content
- For group works, read the question, try to solve it in your head / think of the setup to the solution and think about what steps you need to get to the answer
- Look at the actual answer, if the setup aligns with what you thought of, move on. If not, read through the actual solution and make sure it makes sense, enough that you are confident you will not mess up a similar problem if you saw it.
- Do the same for homeworks!

- Do the same method for going through past exam problems and setting up solutions / comparing to the actual solution using the practice site

### Candus Shi's Advice

- Redo homework and groupwork problems (without looking at the answers, of course). Observe if my approach changed or if I was still getting stuck at the same points as when I first did them. If so, what am I missing and why? If not, hooray! Progress :D
- (In general for math classes) Throughout the quarter, I create a master doc where I write down everything (as concise as possible), organized (ideally sequentially, otherwise poor planning by the instructor) by topic/lecture, and rederive all of the logical steps, including things like:
- Why are we doing this? <Provide explanation; very important to internalize this!>
- Any theorems/proofs/derivations, and redoing them + adding my own commentary to supplement, e.g. properties, intermediate algebra steps.
- Examples, with commentary about what this example is trying to accomplish as an aid to the theoretical concepts.

- Continuing with the above, most of my productive “studying” comes from creating this doc, but sometimes I forget to update it and end up having to do a lot of it closer to the date of an exam; in reality, it’s podcasts 3x speed, all lecture slides/notes pulled up, coffee, jazz, a lot of facepalming, a lot of LaTeX. I also like having this because after the class is over, I have a nice summary of what the class is about and can easily access my notes. This also works better for me because some notes/slides provided by the instructors are hard for me to work with since I was not involved in the thought process of creating them/are confusing/are difficult to read.
- If I get stuck, talking things out with classmates/TAs/professors is more efficient than wallowing in course materials solo! Perhaps this is an incentive to start studying earlier so that one can take advantage of office hours + normal waking hours of friends (though I believe one should study by themselves first).

### Nick Swetlin's Advice

Draw lots of pictures! Be able to draw out loss functions, empirical risk, gradient descent, probability spaces, combinatorics! Don’t just stare at your pictures; ask questions about them! Can I label everything in this picture? What is missing from this diagram? Steal free pictures from the web and gain dozens of new perspectives on a topic! Then, explain all of them to your friends and parents, people who presumably know nothing about data science! You know you’re doing enough when you’re falling asleep at night and your brain all of a sudden says “wait we have to visualize what 5 choose 2 looks like again.”