Potentially useful guides, mostly for students:
Get started in deep learning.
- Go through my research page, and find some interesting problem. Or, define your own problem.
- Read some background material, e.g on Google scholar or wikipedia.
- Revise some mathematics and coding techniques.
- Talk to one of my PhD or MTech students to see if they are already working on a related problem.
- Send me an email with your initial thoughts after the reading and discussion with my students.
- Prepare an agenda and let all participants know.
- Come prepared with some background thinking.
- Seek agreement on decisions or next steps for each agenda item.
- Summarize the meeting's key decisions and next steps for each participant (e.g. in a follow up email).
- Subject line: should be informative to help the recepient assess when to open the email, e.g. "Query about EE201 assignment grading." Never leave it blank. I usually tag my subjects with square brackets, if they belong to one course, e.g. "[EE201] Query about assignment grading."
- Introduction: Start by introducing yourself in one sentence, if the person may not know or remember who you are. E.g. "I am a student in your EE201 class, and was trying to talk to you when you were leaving the lecture hall today." Skip this if the recepient knows you well.
- Purpose: Next, reiterate your purpose in a little more detail. E.g. "I am not clear how the second question was graded."
- Information: Now, organize the information in points, starting from the most important, e.g. "I think that I have arrived at the same solution using a different approach for question 2."
- Extra information: Less important points can come after the most important points, e.g. "Would you be covering the Pentium architecture too?"
- Action item: Tell the recepient what you want them to do, e.g. "Can I meet you tomorrow at noon to go over the assignment?"
Request to work on a project in my group: