It is October already. So, many of you might have started feeling the heat of GATE 2018, especially with the volume of course still left to cover with the added pressure of projects or workload at job. A very alluring option to cope up with this issue is to start byhearting standard results (or formula) and not giving attention to how they came (usually, the mathematics part), as it takes very less time to do so, covers the syllabus faster and most importantly gives us the feeling of being fully-prepared (illusion). We have lot of such standard results in:
- trees (binary, AVL, m-ary, B tree and B+ tree) and graphs,
- computer networks as a whole,
- automata designing (no. of states),
- computer architecture (cache, memory management, disc management and file management),
- digital electronics (combinational and sequential circuits),
- probability and permutations and combinations.
But, is it a wise way to do so? NO, and I will tell you why?
- The fact that GATE examination in CS/IT dates back to 1980s itself nullifies the probability of a direct standard result based question compared to subjects in which GATE has been started recently. Even if such questions comes, 90% of the aspirants will score +1/+2 (full), which means you may have got +1/+2 on paper, but technically you are still at 0 competitively.
- This necessitates the need of questions to be a little deviated from the ideal cases for which those standard results hold which happens in GATE a lot as the question setters already know what kind of loopholes one makes in preparation every-time. Now, the standard results won’t work. But, yes if you learnt the way those results were derived, you can derive the answer for this deviated/ modified/ tweaked case as well. Remember, GATE scrutinizes one’s merit via his/her understanding of the subject, unlike other exams like ISRO, BARC or any other PSUs like BEL, CIL or SBI that tries to gauge one's merit through command over formula based questions (majorly) in their written examination.
- Not only this, this will help you during interviews and written examinations in IITs/ IISc as well, where they very often ask to prove or derive a very commonly known result or some modification of it.
- For example, derive the maximum number of nodes in a m-ary tree of height (h) (root is at height 1). A very handy way of solving this question in an objective type exam would be to take m=2,3,4 and h=5(say) and safely conclude that it will be m^h. But such an approach will fetch you no good repo in an interview or written tests in IITs/ IISc. But, have one learnt during preparation that for a binary tree (2-ary), the formula 2^h comes from a GP of 1,2,4,….,2^(h-1), he/she could easily proceed in the way to get the desired result. Note that, the former was verification while the latter is the proof/ derivation, the terms which are fundamentally different in mathematical premise.
Especially the people, who are taking any coaching as THE BIBLE, have to pay attention to it as the course completion in coaching institutes revolves majorly around giving a brief introduction to the baics and then providing with all set of standard results for advanced concepts, a term called as over-fitting in Machine Learning. Coaching is supposed to be a catalyst and is not a substitute to self-study.
Finally, I would like to address another elephant in the room, which is lack of acceptance. At any point of time if your answer goes wrong, instead of trying to prove it right by bringing a bunch of friends to an online discussion to support you or point others to show what is wrong with your solution with an intention of non-acceptance, realize this for your own good that it is better that your mistake got rectified at this stage and you will not suffer on the exam day. Instead of arguing un-necessarily, focus on LEARNING, which is central to the process of preparation. A problem may have n ways of solving it out of which only a few are correct. So, a complexity analysis of wrong ways will be of O(n) while the analysis of the correct ones will be O(1). This decision to opt for what is efficient as well as beneficial rests with you.
PS: Life is not long enough to learn by committing every mistake on your own. Sometimes, we must optimize our learning algorithm so as to learn from mistakes of others too.
ALL THE BEST !!
HAPPY LEARNING !!