Ugly Duckling theorem

Yes, I know, my last post said that I would be shifting my blog from wordpress to a Jekyll Hugo based blog hosted with Github pages. But I’m procrastinating. Anyway, putting that aside, let’s get to the meat of the post. Check out the Ugly duckling theorem. It’s pretty dope. I encountered it in the introductory lecture slides for the course on Machine Learning I’m doing right now(yikes! ML ooh). Yeah, don’t judge me. I’m not doing ML because EVERYONE is doing ML but because I need to complete a couple of courses from a subset of CS courses to get my CS degree from IITK and it so happens that among those courses, ML is the most theoretical(the other courses are mostly from systems and I (sort of) despise systems courses).

Advertisement

Shifting my site

Hey folks, I thought about this a lot and finally have decided to shift my blog to Jekyll based Github pages due to a couple of reasons:

  1. It’s completely free(until a limit of 1GB). Yeah, wordpress is free but it serves ads and as everybody knows, ads suck.
  2. I bought a domain name for myself – aravindreddy.org and it seems like this is the best time for me to make a switch.
  3. According to multiple online sources, WordPress is outdated and is a pain to maintain, especially when you want to have your own domain name and custom servers.

So, I guess this will be one of my last posts on the WordPress platform. Hopefully, I’ll successfully shift my all my past blog entries to my new Jekyll based, Github hosted site 🙂

GRE craxx Algo

Since I was probably the first person in my batch of ~830 students at IIT Kanpur to have taken the GRE, I keep receiving a barrage of questions from my both my batchmates and juniors regarding how to ‘crack’ the GRE. So, here goes the Algo which I employed to improve my score from 324 to 332 within ~10 days of preparation while studying seriously(sorta) for around 3-4 hours a day:

  1. Download the official Practice Test from the GRE website and take the exam using a timer.
  2. Decide your target score. As far as I know, a bad GRE score hurts your chances of selection but a great score doesn’t improve your chances. As a basic guideline, any score below 320 will hurt your chances of getting into a top 100 university and any score above 330 should be fine (here’s the percentile chart).
  3. The verbal section is the one which requires the most preparation (for most Asian students). So, I’ll focus on that first
  4. There’s no set syllabus for the verbal section. You’ll have to improve your vocab in any way possible.
  5. Starting now, make a habit of looking up the meaning of every unknown word you encounter. This will help you immensely because you’ll learn holistically.
  6. I’ll give a list of books which I used to prepare. But as I stated earlier, there’s no set syllabus and there’s a myriad of books out there which will increase your score. Pick a couple of them and make sure you complete them/revise them regularly before moving onto more books:
    a) Manhattan review: The GRE Complete Guide
    b) GRE Vocab Capacity
    c) GRE Vocab Flash cards app from Magoosh(iOS/Android) – Definitely do the basic ones.
  7. For the Quant section, you can choose to prepare from the Manhattan review(if you think it’ll improve your score). I got a perfect score on the first practice test. So, I didn’t bother to prepare for the Quant section.
  8. The AWA section is a little tricky. I didn’t prepare much but I felt that I wrote one good essay and one decent essay and so I expected a 4.5/6(they give you the average of both the sections, 6 points for each essay) but I only got a score of 4. Since this is subjective, it can get really arbit. I would suggest reading the guidelines in the “Official Guide to the GRE Verbal Reasoning Practice Questions Volume 1”  and then picking topics from the GRE website and practicing with a timer.
  9. On the day before your GRE date, complete the second practice test. This should give you a good estimate of your score.

Scheduling Algorithms in normal life

There exists an interesting correlation between scheduling algorithms in operating systems and scheduling tasks in life. At a fundamental level, humans are no different from operating systems when it comes to scheduling activities. We have a brain and a computer has a processor. We have deadlines, a processor too has them for it’s scheduled jobs.

Turing Award acceptance speech by Dr. Fernando Corbató.

On Building Systems That Will Fail

Goldwasser _/\_

Last semester, I stumbled upon ‘interactive proofs’ and read the word ‘Goldwasser’ as one of the authors of a paper related to interactive proofs(turns out she was one of the inventors of interactive proofs 😛 ). Although there were other authors as well, this particular name struck in my mind. Today, after some arbitrary DFS and BFS’ing through the internet, I read about Prof. Shafi Goldwasser here. I don’t have words to describe the IMMENSE respect I now have for her after reading the article. I really encourage readers to visit that link.

Learning vs Ease of understanding

It is very comfortable to read stuff which we already know about. This is primarily because the brain doesn’t need to form new connections or destroy/modify older connections. There is an inherent inertia in the process of making cells or breaking/modifying brain cells. I think it’s just like physical exercise. The more changes your brain accommodates in less time, the more malleable it remains. Aah, it feels similar to lactose intolerance which I developed over this year. I went through a period of more than 3 months without drinking milk and now it is hard for my body to digest milk. Hmm, our mind is like a sword, which constantly needs to be sharpened, else it will become blunt. So, I think we should actively try to learn things which are HARD to understand for us. It keeps the mind active and sharp. Hmm, interestingly, a large number of physicists in the past have had some sort of inclination towards playing music. I think playing an instrument activates certain parts of the brain which take part in the creation of new ideas. Notable physicists who played instruments include Albert Einstein, Max Born, Werner Heisenberg, Richard Feynman and innumerable others. Learning to play music is SO DAMN HARD for adults that I’m sure it stretches your brain in countless ways.