Well okay, besides having been occupied for the past month, I did very badly this semester so I needed quite a bit of time to get back up to write this post. Nevertheless, I shall try my best to remain objective so as to benefit the readers of this blog.
CS1010E Programming Methodology
This module was taken by Prof Joxan Jaffar. Apparently, he taught fairly well at the start of the semester, only to deteriorate drastically towards the end. The language taught is C.
Weightage
Sit-in labs: 50%
Midterms: 10%
Finals: 40%
This module was basically open-book for every single graded component though I should think that wouldn't help much for a computing module. The main topics covered were control flow, arrays, pointers and structures just like for any other introductory computing module. Sit-in labs were held every odd week (starting from Week 5 onwards) and take-home labs were assigned every other week. If you can do the take-home labs, you might probably consider skipping the lab sessions during the even weeks.
The format of the midterms was as follows: it was a multiple-choice paper comprising of 20 questions; each question has a string of code and you're supposed to interpret it and then choose the option that has the correct output. I actually found it pretty meaningless a paper by the way. But the questions were tricky; sometimes it might just take one semi-colon to make a difference in the solution so exercise extra caution and do not overlook any details. I only got 13/20 which is pretty much below the median I guess. But that is just because I totally neglected one or two topics and there were actually close to 6 questions on those topics so I should think it wouldn't be hard to get above 15. You'll know your midterm results very soon since they'll upload the solutions on the same day itself.
Sit-in labs got progressively harder; there were basically 5 sit-in lab sessions where each subsequent lab was worth 1 more percent than the previous one starting with 8% and ending with 12%. I got A for the first two (but really, anyone can get A for the first one) and B for the remaining ones. That essentially put me in the 50th percentile already.
Finals was just, bad; pointers, structures everywhere. There were two sections in the finals. Section A was like the midterms just that it was no longer MCQ so that you had to write the output in the blank space provided. There were 15 questions on that. Section B was then testing students on their ability to write codes so as to solve the problems presented in the questions. Personally, I thought some of the questions were quite different from those in the sit-in labs but if you're good at computing, it shouldn't make much of a difference. I basically screwed up both sections and I gave up halfway to the extent that I scribbled some nonsense in Section A. That was how badly I wanted to end this and it doesn't really help that it was my first paper.
Result: B-
Yes, first B- though it was kinda expected. My advice is to try and get As for all the labs or at least, 4 out of 5. That can secure you at least a B+, I hope. I actually didn't do much practice for this module except for the take-home labs. In fact, I was quite passive in the sense that I believed practice would only take you so far for a computing module. I am just glad to get over this module which was pretty much useless as C is just obsolete already. If given a choice, I would have loved to take IT1006 but no, CS1010E is a core module for Stats majors.
EC3312 Game Theory and Applications to Economics
This module was taught by Prof Sun Yeneng who was seemingly from the Maths department in FOS. His lectures were so-so as most of the things in the notes were taken from the textbook. I gave up attending them after the first one. Textbook is essential: A Primer in Game Theory by Gibbons.
So there were 4 themes altogether: (1) Static Games of Complete Information, (2) Dynamic Games of Complete Information, (3) Static Games of Incomplete Information and (4) Dynamic Games of Incomplete Information.
Weightage
Tutorial attendance and participation: 5%
Assignments: 10%
Midterms: 35%
Finals: 50%
Basically, attend all tutorials and present once to secure the 5% component, There were 2 assignments, each worth 5%. Assignments were very easy so most people should have no problem securing full marks for them.
The bell curve for midterms was steep; the median was 27/35 and it doesn't help that there are a few PhD students taking this module. So I got only 29 but really, the midterms was actually very easy. The reason why I didn't get above 29 was because there was this particular concept that I didn't clear up before taking the midterms as I started studying only one or two days before the test itself. But the concept itself isn't hard so actually, it shouldn't be a feat to get above 30. Midterms are returned during the next tutorial.
I screwed up the finals especially those questions on Bayesian Nash Equilibrium. Although I could have presented my answers in a more detailed fashion, I don't think it would have helped much as I was pretty weak in that concept itself. The thing about game theory is to really appreciate the big idea. When I was studying for this module, there were many times I had to question myself why this or that has to be done. I just couldn't see the importance in certain things. I mean in reality, most firms aren't going to say, "Ok, so let's formulate our strategies and then work out the Nash equilibrium." It just defeats the purpose of game theory. I took this module actually to train my logic and learn to understand from the perspective of a firm. Unfortunately, I thought there was way too much focus on theoretical concepts so that I didn't actually reap much from this module.
Result: B+
B+ was not shocking given that there were probably only a little over 50 people taking this module and that this module didn't really come as intuitive to me. My advice is that divert more attention to the topics covered after midterms as those will be the main focus of the finals. Well honestly (and I'm not saying this cos' I didn't do well), this module is close to useless unless you want to go on to higher-level modules or do your thesis on something related to game theory. I mean the emphasis on the theories of Nash Equilibrium was just way too much such that I don't see the realism in this module.
EC3333 Financial Economics I
This module was taught by Prof Lu Jingfeng whose English may be a bit hard to comprehend (I seldom comment on such things but if I have to say it, that just means something). But then again, if you don't go for lectures...
And so this was one module I really disliked and for the first time, I actually regretted taking a module. I thought nothing could get worse than micro in terms of the level of interest but this was just many times worse than micro. Aside from the fact that everything was being thrown at you without any avenues to enhance your understanding (I tried to look for a book that derives all the theories and formulas but to no avail, neither was asking the Prof a solution as well), this module was way too technical for my liking. So it's just formulas and graphs and formulas again.
Topics covered are Optimal Risky Portfolios, CAPM, APT, Bonds and Options. I personally think the textbook was bad (no derivations) but still necessary for this module since the notes are mainly slides created out of the content in the textbook: Investments and Portfolio Management by Bodie, Kane and Marcus, 9th Edition. Note: Investments by Bodie, Kane and Marcus, 9th Edition is almost identical and can be used as a substitute.
Weightage
Tutorial Attendance and Participation: 15%
Assignments: 15%
Midterms: 30%
Finals: 40%
The first component is supposedly easy to get, just attend all tutorials and participate three or four times in class. Similarly, the second component is supposedly easy to get as well but unfortunately, I did not get full marks for the last two assignments. Basically, how it works is that each student gets allocated 3 assignments at random so some may get the ones in the earlier weeks and some may get a mixture of those in the earlier and later weeks. Well, the impression it gives is that it might be somewhat unfair since those who got allocated the tutorials covered in the earlier weeks (especially before the midterms) have it easier. But really, I think the assignments are all easy and given enough understanding, one should be able to score full marks for all of them.
The midterms was disastrous for me. I got a mere 21/30 when the average was probably around 25.5 to 26. Reason: This exact same question from the practice midterms came out and I screwed it up. Yes I did not practise the midterms but aside from that, I think even if I did not, I should have been more careful and have a deeper understanding of the concepts involved. Then there was another question in which I got the correct answer at first but having been too paranoid, I decided to act smart and write something else so I got that wrong too. In fact, there was only one 'differentiator' question which oh, I happened to answer it wrongly as well. That, I can only blame myself for not understanding the earlier chapters thoroughly. In fact, I think the midterms is really doable and it wouldn't have been hard to get at least 26 or something along that line. Anyway, midterm scores will be uploaded onto IVLE once grading has been done.
Finals. Finals was apparently easy since I got a B and I thought I did pretty well. So it's either I underestimated the cohort (doesn't help that there are a few DDP students taking this module) or I overestimated myself. I think it's a mixture of both. Proofs can be easily found in the textbook and yes, contrary to what I thought, many people actually read them. Other than that, it's all just simple calculations and manipulation for one question.
Result: B
ST2132 Mathematical Statistics
Well, the first thing I must say is that as intimidating as Prof Lim Chinghway looks, he's one heck of a good lecturer. He's patient, willing to help, explains concepts really clearly and would even repeat if the class did not catch it. Possibly almost everything you would want to see in your lecturer.
The fact is that this module is very easy compared to ST2131. Like I said, Prof Lim toned it down a lot so that it transformed from a killer module to a CAP-puller for most people I guess.
The big topics covered included Simple Random Sampling, Parameter Estimation (MOM and MLE), Fisher Information, Efficiency, Sufficiency, Hypothesis Testing, Generalized Likelihood Ratio Test and the Comparison of Two Samples. Textbook isn't necessary at all since the notes are sufficient and tutorial questions are from there as well but you could still get it if you want: Mathematical Statistics and Data Analysis by John Rice, 3rd Edition.
Weightage
Tutorial Attendance: 5%
Tutorial Participation: 5%
Assignments: 20%
Midterms: 20% (graded on completeness and also, correctness from Week 2 or 3 onwards cos' they found a grader)
Finals: 50%
Bonus: 5% (for participation either in class or on IVLE Forum)
The first 3 components are free marks so there's no need to talk about them (because you are actually allowed to make changes to your tutorial answers when your tutorial mates are presenting their answers).
I screwed up midterms as usual, got only 24/40. Average was 22 or something. The questions were easy but one of them was pretty unprecedented. But on overall, I must say the Semester 2 paper was harder than Semester 1 paper (and much less people take ST2132 in Semester 2). Even so, it was still doable. Highest was 38/40 by the way. So the S.D. is pretty high. The midterm score was uploaded onto gradebook before the scripts were returned to students during tutorial.
Finals was easy as well. Know your concepts well and trust me on this, copy down all the probability mass or density functions onto your cheatsheet. You'll need it during the exam. With that, you can easily get a decent grade provided you know what you are doing.
Result: A-
Probably got saved by the finals.
ST3131 Regression Analysis
Personally, this is one of the hardest Stats modules I have taken. The level of understanding involved in this module on top of the derivations is not trivial. Oh plus the bell curve is very steep with lots of Maths majors screwing it up. From what I heard, this module used to be on the same level as ST1131. However, since Prof Anthony Kuk took over, it became quite difficult a module.
Coupled with procrastination, I was always lagging behind lectures (I don't go for lectures partly cos it's 8am and partly cos going for lectures isn't gonna make much of a difference). I never attended tutorials as well since the tutor is atrocious as everyone has made him out to be.
Topics covered ranged from 1-factor ANOVA, 2-factor ANOVA to simple, multiple, subset regression, residuals, outliers and the combination of ANOVA with regression. There was no compulsory textbook for this module. All were just references but the main one was Introduction to Regression Analysis by Montgomery, Peck and Vining, 5th Edition.
Weightage
Assignments: 20% (2 assignments worth 10% each)
Midterms: 20%
Finals: 60%
Most people got full marks for assignments which mainly made use of R (I did not though). R isn't tested in finals by the way. Midterms was just plugging in formula. Well, the thing about midterms was that you didn't have to understand what you are doing, you just have to know what to use. Average was around 30/40. The midterm score was uploaded onto gradebook before the scripts were returned to students during lecture.
Finals was the hard one for me or probably for many people. So nothing in the notes appeared and they were all out-of-context questions. This just boils down to how well you understand the concepts and yes, because I didn't, I screwed the finals up badly. But the finals was really typical of an open-book exam where you had to think on the spot and manipulate some stuff. Time wasn't a constraint though.
Result: B
No doubt finals is the determining factor since I got full marks for midterms but at the same time, no doubt that this module is important. My advice is be consistent, clarify doubts immediately, don't ever snowball anything.
