ST4240: Data Mining
Lecturer: Alexandre Thiery
This module can be used for specialisation and it is a very popular module due to the whole "big data" trend recently. I found this module to be extremely useful as it tackles real-life scenarios. As statistics students, we are definitely going to face loads of data in the future. This module will introduce you to some common predictive methods used for large datasets and provide ample practice for a good foundation.
After taking ST3233 under the same lecturer last semester, I already knew what to expect for this module. As always, the lecturer's notes are very simple with not much information, but he will always write notes during the class and upload them onto IVLE. The tutorials consist mostly of coding questions, especially in the later few chapters when we start learning all the different predictive methods. Webcast is provided for those who did not attend the lecture and surprisingly, the lecturer is actually fine with us not attending LOL. Ok, I admitted I skipped a few lectures due to my unsatisfactory timetable but at least I attended more than 50% :)
Workload wise, they all come in the second half of the semester. There are 3 group assignments in total and all revolve around coding. You will receive large datasets and have to predict the outcomes. 1 of the assignment lasted the whole semester. We had to predict some values and upload them onto this website called Kaggle, where there is a leaderboard to see how well our prediction fares compared to other groups. The midterm and final were both closed book and no cheatsheet, which is the lecturer's style. No surprise there. Since coding cannot be tested in a pen-and-paper test, expect a bunch of questions testing your conceptual knowledge. I think I did really badly for that. Anyway, I would strongly recommend all statistics students to take this module as I personally feel that we really need some practical knowledge to compete with other grads from computing, business analytics, etc. A lot of our stats modules are theoretical and not really useful for work.
ST4245: Statistical Methods for Finance
Lecturer: Xia Yingcun
This module is also popular because of its name. Like, come on, there is "Finance" in the module name. Obviously we would try to clamour for it. Ok, so this module covers quite a broad range of stuff. There is some portfolio theory which was also taught in MA3269 and time series which was taught in ST3233 and some other econs modules. I would say this module might be useful as it does cover a lot of topics but statistics is built on a lot of theories (read: assumptions) and in real life, you cannot expect all your data to be so nice...
This lecturer is ok I guess. He is my worst lecturer for this semester but I had good lecturers for my other modules so I would rank him average. It might be a little hard to understand his accent and sometimes I have no idea what he is writing on the notes but I just copy anyway. If you did not catch what he is saying, you can always watch the webcast. His notes are very simplistic (read: no design) and sometimes not really organized well. I think that as long as you pay attention during his lectures and practice his tutorials plus some other questions, you should score decently for this module.
On to the workload part. So every week there is a tutorial BUT you have to do them and submit. In groups of 10. That is like the biggest group I worked with in my entire life. Thankfully, his tutorials have a lot of repeated questions from last year so a lot of groups just copy those answers. NOT thankfully, he also repeated questions for his midterms. He repeated questions in an open book test. So people with access to past year answers could just copy and get like... nearly 60/100 marks free. Wtf. Some people complained after the test but all he did was to reduce the weightage of the midterm. Well, at least the open book finals were fresh questions. Unless there were repeated questions that no one told me about. The horror.
EC2204: Financial Accounting for Economists
Lecturer: Chua Yeow Hwee
Hmm I actually did not plan on taking this module, but my appeal for the third core module failed so I settled for an elective instead. As the module name states, this is an economics module but it covers basic accounting and also precludes the accouting module over at the business school. You will learn about the basic stuff in all those financial statements and know how to record and analyse the financial statements. Apparently this module used to be offered a long long time ago and it was only revived this year. So I guess I was part of the new guinea pig batch?
The lecturer is enthusiastic and good. He teaches through examples which suit me just fine and he frequently mentions recent news articles to show us how relevant this module is. Webcast is also provided. His notes are really thick because there are tons of stuff to learn and he puts practically everything in his notes. You can see the effort put in by the lecturer as the notes are quite time-consuming to make. The tutorials are also all taught by him (whoa one man show) so you can see his friendly face twice a week.
Workload wise, you should do the tutorials before attending as you might be asked to answer the questions. But it is ok if you really cannot answer. There is a group project of 3-4 members which make up a large percentage of your grade. The group has to choose any company from the SGX website and analyse their financial performance over the past 3 years, then submit a report and do a group presentation. Midterm and final were closed book. Midterm had some MCQ questions but final did not and you will also have to prepare some financial statements from scratch during the final. I found this module useful and this is a good alternative if you do not wish to compete with the business students in their accounting module. It is always good to have some basic accounting knowledge so you know where money flows to in companies.
LAK2201: Korean 2
Lecturer: Chi Seo Won
Lucky me. I managed to appeal into the module (probably because I gatecrashed the first lesson) and had the same lecturer as last semester too! Korean 2 is like a level up from Korean 1. Expect a slightly faster pace with more grammar, more vocabulary, more work. Other than that, the format is basically the same as Korean 1. Please ensure that you have a good foundation from Korean 1 before taking this module or you might die.
Well, having the same lecturer is good since I already knew how interesting she can be. As usual, she can be a little dramatic at times but still teaches really well. I wouldn't really consider her a lecturer. She is more like a teacher, those kind that you have in kindergarden or primary school, always energetic and enthusiatic. Lessons are always fun and very interactive. You will definitely need to speak up at least twice every lesson and if you come early, you will probably end up having a korean conversation with her.
Workload wise, it is more than Korean 1. There are a few written essay assignments with longer word requirements that are scattered throughout the semester. Also, there are 2 oral tests (1 reading a passage and 1 face-to-face) as well as a midterm and final. The new workload comes in the form of a pairwork whereby you have to partner someone and film (in one take) the two of you having a conversation (write your own script) based on the given topic for about 1 min 30 secs. Other than that, everything is about the same as Korean 1, so you should be used to the testing style and everything already. Expect stiffer competition since those who continue on to Korean 2 are those who did well for Korean 1. 화이팅!
GEK1505/GEH1036: Living with Mathematics
Lecturer: Lou Jiann-Hua
This module covers a very broad range of topics that you may encounter in daily life: counting (series and permutations and combinations), graphing (basic graph theory), clocking (periods and modulus), coding (not the computer programming but the secret message kind of coding), enciphering (create and decode secret messages) and chancing (probability). Expect a whole bunch of fake noobs from maths, stats and other similar majors coming in to take this module. And therefore, expect very steep bell curve.
The lecturer has taught this module a few times already. He is quite old and speaks slowly, but he explains the stuff well. The lectures are from 6-8pm and webcast is available so I just skipped all the lectures and watch the webcasts at 1.5x speed. If you are smart and hate watching webcast, it is fine too since the notes are self-explanatory and full of examples. It is possible to just read the notes, do your cheatsheet and waltz into the exam hall since the module difficulty is not high. But I would still try to encourage you to support the lecturer and either attend the lectures or watch the webcasts. :)
There is practically zero workload for this awesome module. The midterm is 30% and the final is 70%. 1 piece of cheatsheet is allowed for both of them so just copy the necessary stuff and maybe a few examples. Tutorial attendance is taken but not counted towards your grade. My tutor was pretty ok at going through the questions so I just went for the tutorials to support her. There is nothing much to do for this module. Just go through the tutorial questions and understand them for a decent grade. There are also a whole bunch of past year papers to try and you will realise the questions are about the same. So if you are good at math and logic stuff, feel free to take this module to lighten your overall workload.
