jOSiAh's ^_~ Haven

Semester 2 has ended about a month ago (well, for most students anyway...). Most undergraduates are already on their summer break, and many Malaysians I know have already gone back to Malaysia to sweat under the hot sun (^,^)

Personally, I felt that my second semester was more relaxed than my first. Compared to the first semester where summative coursework accounts for 100% in some of the modules, it is only worth 40% in semester 2 (with 60% being exams) for all modules. And we Malaysians are trained to be vigorous, robust exam machines, so that's not much of a problem. And the fact that the modules are more specialised this time around, there are fewer weird, fuzzy essay-type courseworks (which I really hate writing), but more of the problem-solving, practical kind (yippee!!). That made my life much easier.

Anyway, a recap for what I have learnt during the second semester...

Vision: A computer vision course. Set to be a very practical course rather than a theoretical course, with lots of classic materials as well as very current and up-to-date techniques. Topics covered include image segmentation, edge detection, salient points detection, object recognition (simple shapes, variable shapes, using templates/patches), stereo reconstruction, bayesian networks. Courseworks include texture segmentation and classification of cars and motorbikes. Ended up learning things which I would have never even thought about learning if I were to do my Masters by research in MMU. Of course, it ended with a tough open notes exam where you end up thinking during the exam more than writing your answers (>.<)

Language: A natural language processing course. Since the school's expertise is more towards statistical NLP, that's what I've been learning most of the time. Topics covered: Corpus, information retrieval, information extraction, counting, data sparseness, n-gram modelling, and of course classical NLP stuff: POS tagging, grammar, parsing, ambiguity etc. Courseworks: give a student presentation on a chosen topic, and do some unsupervised learning to solve an NLP problem. And again, ended with an exam, a closed book one this time.

Bio-inspired computing: Main topics include swarm intelligence (Ant Colony Optimisation etc), collective behaviour, neural networks (neurons, ANN, RNN, DNN, Hopfield Networks, Kohonen/SoM etc.), evolutionary computing (Genetic Algorithm), co-evolutionary computing. Courseworks: do some "mind reading", try to play around with genetic algorithms, and finally a presentation on a bio-inspired topic. Ended with yet another closed book exam. Oh and I have also found an interesting "practical" application of SoMs whilst preparing for my exams. Never thought that preparing for exams could be so fun! (^,^).

Computer Security: It was either this elective or another elective called "Architectures for Complex IT Systems". Fortunately, I chose this one (the other had no exams, but had 2 essay-based courseworks (*shivers*) with 2 presentations, and a 20% test... oh no way... that's too much work for me). Unexpectedly, I seemed to have learnt a lot from this course. We did not go into as much details as the undergraduates did (*whew*), as this module was meant as a generalists module. Topics include malicious software (trojans! worms!), software vulnerabilities (SQL injection, cross-site scripting, buffer overruns etc.), threat modelling, cryptography, authentication, network security, intrusion detection, and computer forensics. Courseworks include "hacking" my lecturer's little offline "bank" system, and writing a report on a security topic (I did one on Honeypots). And of course, exams (>.<).

I personally think that this semester is much fruitful than the previous semester since it is more geared towards my course (Cognitive Systems) compared to the more "general" stuff in the first semester.


So now, during the summer break, I will be working full time on my project which I have already been working on since the whole semester. What fun! (>.>)