So, I clicked “submit” on the last question of the final exam a couple of hours ago, and I'm satisfied to note I have an overall 92% grade. But was it a good course?
Ooh, let's rewind a bit. This course is supposed to be many things, an introduction to the world of genomic, or precision, medicine for the layman as well as for medical students. In eight short weeks, we get maybe a dozen different people giving short lectures about their particular field, all in relation with genomic medicine. The lectures are arranged in four themes, which are broadly the clinic, the lab, business, and ethics. Dr Haddad, the main instructor, introduces everybody and conducts Q&A sessions, in order for the course not to feel too disjointed.
So, a success? Partly. The production is good, although maybe they've overdone it a bit. Before each video segment we have a full page of text explaining the pedagogical objectives of the segment. Big instructions in bold remind us to use the navigation bar to, erm, navigate the course. And so on. It feels somewhat dumbed-down.
The course content is, obviously, varied. Generally I liked the clinic theme and was bored with the other three (the lab one was very introductory). Although they emphasize that medical students may be watching the videos, in truth − I guess medical students have better things to do. The overall level of the course is very, very basic. Apart from a tidbit here and there (I wasn't aware of fluorescent in-situ hybridization), I can't say I have learned much during the course, and only stuck with it because it's summer and I had nothing better to do.
Hmm… I may sound too harsh. Let's say that if you approach this course as you would, say, one from MIT, then you'll be disappointed. Which is not to say there isn't merit; the team have made a good attempt at surveying the landscape of genomic medicine, primarily from doctors' points of view, and deliver something akin to an Internet-era pop-science book, and a pretty decent one at that.
So, would I recommend this course? Not generally, not to people with a utilitarian view of MOOCs, who measure success in skills and knowledge acquisition. But to specific people, maybe to people who're simply curious, who've heard about sequencing the genome, maybe about the recent judicial issues surrounding 23andMe, and who are willing to spend a couple of hours a week to find out more about it, yeah. Definitely.
Wednesday, July 30, 2014
Saturday, July 26, 2014
Midyear MOOC review
We're about halfway through the year - or at least, we're halfway through the summer holidays in the northern hemisphere, which in academic terms means we're at the midpoint of the year although we're 7/12th of the way to Christmas. Anyway; now's a good time for a review.
2013 was my first contact to MOOCland; 2014 is my first year of going full blast with Education 2.0. So far I've completed 14 courses this year, only one of them a carryover from 2013. I'm registered for an additional 13, and while I am almost certain to drop some, I'm fairly certain I won't stop there. As a comparison, last year I only completed 4 courses (6 if counting the two I audited); I've certainly gained some momentum there.
That's not the only difference. Not only did I get fully into this MOOC thing, I also have moved a bit from learning for learning's sake, and started focusing on a sort of pathway that may lead eventually to my career changing tracks. Which is funny, in a way, because I got onto the bandwagon to learn some economics theory, something that's never going to feature significantly in my work. Rather, I'm now gathering most of my efforts towards building a scientific curriculum centered around the general theme of computational biology. While my sense of fun is still my main driver in selecting courses, I've also come to pick up subjects in a utilitarian fashion, for instance taking all three Statistics courses from UC Berkeley (and the Analytics Edge from MIT) because I felt the need to brush up on the subject in order to better tackle the biology field.
Anyway; a half-year in review.
JANUARY saw the end of Harvard's slightly overproduced MCB80x - Fundamentals of Neuroscience. In retrospect, the course was better than what it felt at the time; I only wish there had been a quicker pace to it, and less time spent on cartoons and more on "test yourself" type exercises. (The instructor hates testing and grading − for reasons I can empathize with, while not really agreeing with − so this course has some virtual labs and a final exam, and that's it. No intermediate assignments, homework, etc. to fix things into the students' minds. Combined with a short-chapter-every-fortnight pace, that makes for a course that's very low-intensity and so, hard to commit to memory.)
It was also the start of CalTech's Principles of Microeconomics with Calculus, a very intensive course, very challenging, but quite rewarding too. I took it because, hell, I got into online ed to learn some econ, I wanted to swallow the whole pill. While the course ended up convincing me I didn't want to study economics for a livelihood (fat chance of me ending up that way even if I wanted to anyway) it managed to burn some basic economic concepts (supply, demand, monopoly, oligopoly, externalities, Pigouvian taxation, optimality, etc.) into my mind, which proves invaluable to my world-view generally. I haven't taken more economics courses since; when I get the motivation for it, there's the archived development economics course from MIT's Duflo and Bannerjee (both superstar economists in their own right) waiting for me.
In MARCH I started doing some serious MOOCing. It was the start of Berkeley's Stats course (which is both great and too easy − the contrast with their Californian neighbours at CalTech was terrible: CalTech's Rangel would have gone through the whole 15-week stats curriculum in, like, 5 weeks), that I took because I knew I was too ignorant in the way of statisticians for my own good; at the same time I started Rice's Fundamentals of Immunology (to keep doing some biology) and MIT's Analytics Edge (a very nice hands-on, reality-based complement to Berkeley's abstract, theoretical stats).
I also picked up MongoDB's DBA course at that time. It's a MOOC, running the edX software and produced in a very similar way to any of MIT's or Harvard's, but not offered by a university and very much focused on practical skills on a specific product, so I don't know if it really counts. I find it sociologically interesting though: the folks taking the course (or at least, those that were vocal on the forums) are very different from the MOOC-taking population on Coursera or edX. Let's be euphemistic and say it's a different mindset (in less charitable terms, a lot of people are only there to get a certificate to stick on their CVs and couldn't care less about the subject matter, or are plain stupid, or are whiny kids. Or all three at the same time.) Still, I admire the instructors' patience and I did learn a lot of useful stuff; never mind the forums.
Still in March, I started Copenhagen's Diabetes Challenge course (because my 4-year-old son has Type I diabetes and I'm very interested in the bio/healthcare sciences), which proved unequal but pretty interesting, Peking University's Bioinformatics course (which suffered from being dubbed in English by non-native speakers; it's awful to say that, but the delivery really hurt. Another big, big problem with that course is, well, honestly, how can you have an algorithmics course where you don't write a single line of code?) and ANU's first Astrophysics module (for no better reason than to get bragging rights, as in: "I studied under a Nobel prize winner"), which proved a blast (no pun on bioinformatics intended).
That was pretty much it (6 or 7 concurrent courses are pretty much my absolute limit) until the end of APRIL, when I started the eagerly-awaited Epigenetics course from Melbourne University, which picked up pretty much where MIT's 7.00x had left, genetics-wise, and was very, very good. Surprisingly difficult, because I don't know how to read scientific papers, really; but very stimulating. Around that time, I tried Mount Sinai's Introduction to Systems Biology and Harvard's Data Analysis for Genomics, but dropped both: they were too advanced, I was too tired. Instead, I refocused on finishing what I'd started (picking up Stat 2.2x 3/5th of the way through, and still managing an overall 65%) and cruised at 5-6 simultaneous courses until MAY, when a lot of courses ended. In retrospect, it's been my most fruitful month ever, as I picked up certificates for Analytics, Diabetes, Astrophysics, and Statistics (part 2).
May has also been the month of my biggest disappointment in MOOCland: while MIT's courses have generally been head and shoulders above the rest of the crowd, the Social Physics "buy-my-book" ad was a downright scam. I still don't understand how or why MIT and edX have let this pass, but hey, let's not take it against them and, well, be more selective in the future. It's a good reminder that the best can rub shoulders with the worst, I suppose, and that we should not take quality for granted.
On a similar note, I've been mostly an MIT/edX fanboy, because that's how I got into this MOOC thing, you know? But while I still prefer the edX platform overall (because of the more linear flow and richer interactive grading options, and despite the rubbish forums) I've gotten somewhat neutral. There are some great courses on Coursera too (starting with Melbourne's Epigenetics), and while the overall system feels more rigid, it's also less prone to bugs and delivers a consistently good experience − and the courses, well, they're often from less prestigious universities and tend to have less whizz-bang than the big kids at MIT-Harvard-Berkeley, but they're pretty good nonetheless. One just has to be more selective − and not hesitate to trial courses and drop the ones that don't fit.
JUNE has seen the end of Epigenetics and the start of a small bunch of courses: Georgetown's Genomic Medicine (which I feel ambivalent about: it's more an outreach program than an actual course. There are good things there, but no deep science − it's more of an extended documentary about the impact of genomic technology on the practice of medicine today. If nothing else, it reminded me of the difference between biological science and medicine, a difference that I didn't perceive fully twenty years ago, the reason why I opted for maths/physics rather than biology in my formal education), MIT's 7.QBW (a great, though frustrating, glimpse of what computational biology can be, which motivated me to try again Mt Sinai's Intro to Systems Biology in September), ANU's second Astrophysics course, this time about Exoplanets (more because it's fun and stimulating than because the co-instructor has a Nobel, this time), and the last part of Berkeley's Stats program.
At the end of JULY, 7.QBW is finished, Genomic Medicine is in its dying throes, Exoplanets is rolling along, and I've just started U. Illinois' Emergence of Life, which feels… I don't know, haphazard? Anyway, it's as good an introduction to evolutionary biology as I'll get in the summer − MOOCs are going slow until September.
So… What next? Well, I'm registered for a whole bunch of courses. First, I'm trying out the UK's own FutureLearn platform to learn about the Scottish independence referendum (my grandfather was Scottish, so I feel kind of romantically attached to the land of Ayes and Scotch, although I've only ever been there as a tourist). The course straddles the referendum itself, from Aug 25th to the end of September, so we'll get to learn about the issues then about the aftermath (although it's pretty clear the No will win). I don't expect this course to take up much of my time.
Much more seriously, come SEPTEMBER I'll be taking:
In OCTOBER most of these courses will still be running (all except Scottish Independence and Explore Stats) but I'll still be starting Delft's Engineering for Bio-based products, because I feel it's my continental duty to pick up some European courses (what do you mean, I'm not convincing?) and because it's somewhat intriguing. Also, by the end of the month starts the second part of Rice's Fundamentals of Immunology, that I'm pretty committed to seeing through (I've done the first part and I hate leaving things halfway done).
Sometime in Q3 2014 (whatever that means) starts Harvard's Muscoloskeletal Anatomy − a good complement to Duke's Physiology, with virtual labs (including dissections). Also, if I see the Systems Biology through, then I'll probably embark on Experimental Methods in Systems Biology and the rest of the specialization.
So, that promises to be quite an eventful second half of the year. We'll see how it turns out!
On a side note, I'm starting to wonder about turning all this knowledge into an actual degree and a career change. There seems to be options with the CNAM (France's continuing education university-like institution, aimed at working professionals) so I'll be contacting them in September to study things through. Paradoxically enough, if I do end up taking evening classes there, it'll be because of MOOCs − but it'll also mean I can't take MOOCs anymore for time reasons. Oh well, that's all very hypothetical. We'll see!
2013 was my first contact to MOOCland; 2014 is my first year of going full blast with Education 2.0. So far I've completed 14 courses this year, only one of them a carryover from 2013. I'm registered for an additional 13, and while I am almost certain to drop some, I'm fairly certain I won't stop there. As a comparison, last year I only completed 4 courses (6 if counting the two I audited); I've certainly gained some momentum there.
That's not the only difference. Not only did I get fully into this MOOC thing, I also have moved a bit from learning for learning's sake, and started focusing on a sort of pathway that may lead eventually to my career changing tracks. Which is funny, in a way, because I got onto the bandwagon to learn some economics theory, something that's never going to feature significantly in my work. Rather, I'm now gathering most of my efforts towards building a scientific curriculum centered around the general theme of computational biology. While my sense of fun is still my main driver in selecting courses, I've also come to pick up subjects in a utilitarian fashion, for instance taking all three Statistics courses from UC Berkeley (and the Analytics Edge from MIT) because I felt the need to brush up on the subject in order to better tackle the biology field.
Anyway; a half-year in review.
JANUARY saw the end of Harvard's slightly overproduced MCB80x - Fundamentals of Neuroscience. In retrospect, the course was better than what it felt at the time; I only wish there had been a quicker pace to it, and less time spent on cartoons and more on "test yourself" type exercises. (The instructor hates testing and grading − for reasons I can empathize with, while not really agreeing with − so this course has some virtual labs and a final exam, and that's it. No intermediate assignments, homework, etc. to fix things into the students' minds. Combined with a short-chapter-every-fortnight pace, that makes for a course that's very low-intensity and so, hard to commit to memory.)
It was also the start of CalTech's Principles of Microeconomics with Calculus, a very intensive course, very challenging, but quite rewarding too. I took it because, hell, I got into online ed to learn some econ, I wanted to swallow the whole pill. While the course ended up convincing me I didn't want to study economics for a livelihood (fat chance of me ending up that way even if I wanted to anyway) it managed to burn some basic economic concepts (supply, demand, monopoly, oligopoly, externalities, Pigouvian taxation, optimality, etc.) into my mind, which proves invaluable to my world-view generally. I haven't taken more economics courses since; when I get the motivation for it, there's the archived development economics course from MIT's Duflo and Bannerjee (both superstar economists in their own right) waiting for me.
In MARCH I started doing some serious MOOCing. It was the start of Berkeley's Stats course (which is both great and too easy − the contrast with their Californian neighbours at CalTech was terrible: CalTech's Rangel would have gone through the whole 15-week stats curriculum in, like, 5 weeks), that I took because I knew I was too ignorant in the way of statisticians for my own good; at the same time I started Rice's Fundamentals of Immunology (to keep doing some biology) and MIT's Analytics Edge (a very nice hands-on, reality-based complement to Berkeley's abstract, theoretical stats).
I also picked up MongoDB's DBA course at that time. It's a MOOC, running the edX software and produced in a very similar way to any of MIT's or Harvard's, but not offered by a university and very much focused on practical skills on a specific product, so I don't know if it really counts. I find it sociologically interesting though: the folks taking the course (or at least, those that were vocal on the forums) are very different from the MOOC-taking population on Coursera or edX. Let's be euphemistic and say it's a different mindset (in less charitable terms, a lot of people are only there to get a certificate to stick on their CVs and couldn't care less about the subject matter, or are plain stupid, or are whiny kids. Or all three at the same time.) Still, I admire the instructors' patience and I did learn a lot of useful stuff; never mind the forums.
Still in March, I started Copenhagen's Diabetes Challenge course (because my 4-year-old son has Type I diabetes and I'm very interested in the bio/healthcare sciences), which proved unequal but pretty interesting, Peking University's Bioinformatics course (which suffered from being dubbed in English by non-native speakers; it's awful to say that, but the delivery really hurt. Another big, big problem with that course is, well, honestly, how can you have an algorithmics course where you don't write a single line of code?) and ANU's first Astrophysics module (for no better reason than to get bragging rights, as in: "I studied under a Nobel prize winner"), which proved a blast (no pun on bioinformatics intended).
That was pretty much it (6 or 7 concurrent courses are pretty much my absolute limit) until the end of APRIL, when I started the eagerly-awaited Epigenetics course from Melbourne University, which picked up pretty much where MIT's 7.00x had left, genetics-wise, and was very, very good. Surprisingly difficult, because I don't know how to read scientific papers, really; but very stimulating. Around that time, I tried Mount Sinai's Introduction to Systems Biology and Harvard's Data Analysis for Genomics, but dropped both: they were too advanced, I was too tired. Instead, I refocused on finishing what I'd started (picking up Stat 2.2x 3/5th of the way through, and still managing an overall 65%) and cruised at 5-6 simultaneous courses until MAY, when a lot of courses ended. In retrospect, it's been my most fruitful month ever, as I picked up certificates for Analytics, Diabetes, Astrophysics, and Statistics (part 2).
May has also been the month of my biggest disappointment in MOOCland: while MIT's courses have generally been head and shoulders above the rest of the crowd, the Social Physics "buy-my-book" ad was a downright scam. I still don't understand how or why MIT and edX have let this pass, but hey, let's not take it against them and, well, be more selective in the future. It's a good reminder that the best can rub shoulders with the worst, I suppose, and that we should not take quality for granted.
On a similar note, I've been mostly an MIT/edX fanboy, because that's how I got into this MOOC thing, you know? But while I still prefer the edX platform overall (because of the more linear flow and richer interactive grading options, and despite the rubbish forums) I've gotten somewhat neutral. There are some great courses on Coursera too (starting with Melbourne's Epigenetics), and while the overall system feels more rigid, it's also less prone to bugs and delivers a consistently good experience − and the courses, well, they're often from less prestigious universities and tend to have less whizz-bang than the big kids at MIT-Harvard-Berkeley, but they're pretty good nonetheless. One just has to be more selective − and not hesitate to trial courses and drop the ones that don't fit.
JUNE has seen the end of Epigenetics and the start of a small bunch of courses: Georgetown's Genomic Medicine (which I feel ambivalent about: it's more an outreach program than an actual course. There are good things there, but no deep science − it's more of an extended documentary about the impact of genomic technology on the practice of medicine today. If nothing else, it reminded me of the difference between biological science and medicine, a difference that I didn't perceive fully twenty years ago, the reason why I opted for maths/physics rather than biology in my formal education), MIT's 7.QBW (a great, though frustrating, glimpse of what computational biology can be, which motivated me to try again Mt Sinai's Intro to Systems Biology in September), ANU's second Astrophysics course, this time about Exoplanets (more because it's fun and stimulating than because the co-instructor has a Nobel, this time), and the last part of Berkeley's Stats program.
At the end of JULY, 7.QBW is finished, Genomic Medicine is in its dying throes, Exoplanets is rolling along, and I've just started U. Illinois' Emergence of Life, which feels… I don't know, haphazard? Anyway, it's as good an introduction to evolutionary biology as I'll get in the summer − MOOCs are going slow until September.
So… What next? Well, I'm registered for a whole bunch of courses. First, I'm trying out the UK's own FutureLearn platform to learn about the Scottish independence referendum (my grandfather was Scottish, so I feel kind of romantically attached to the land of Ayes and Scotch, although I've only ever been there as a tourist). The course straddles the referendum itself, from Aug 25th to the end of September, so we'll get to learn about the issues then about the aftermath (although it's pretty clear the No will win). I don't expect this course to take up much of my time.
Much more seriously, come SEPTEMBER I'll be taking:
- Data Analysis and Inference from Duke University - a more practical (with R) overview of statistics. Not very high on the priority list, more like a way to keep the stats knowledge warm.
- Physiology from Duke University also - something I intend to put quite a lot of hours into. Medicine without the practice-of-medicine angle, just the thing for me. (For the summer I've downloaded a 1500-page textbook on Physiology from OpenStax, in order to get a heads-up).
- Exploring Neural Data from Brown. A follow-up on MIT's Quantitative Biology, this time focused on neurology and Python. I have high hopes for this one.
- Introduction to Systems Biology from Mount Sinai. Hopefully this time I'll be able to follow the professor along; I'm still convinced the subject is very interesting, despite the professor being, hmm, clearly a much better scientist than teacher.
- Explore Statistics with R, from Karolinska Institutet. Just because I want to hear Swedish accents again − and okay, because it claims to teach where to get good healthcare-related data too. I guess I'll skip the parts about learning R.
- Introduction to Dinosaur Paleobiology (Dino101) from Alberta University. I'm a 37-year-old kid, so what? Anyway, everybody says the course is enjoyable but low-intensity, which is fine with me, what with all the other courses at the same time.
In OCTOBER most of these courses will still be running (all except Scottish Independence and Explore Stats) but I'll still be starting Delft's Engineering for Bio-based products, because I feel it's my continental duty to pick up some European courses (what do you mean, I'm not convincing?) and because it's somewhat intriguing. Also, by the end of the month starts the second part of Rice's Fundamentals of Immunology, that I'm pretty committed to seeing through (I've done the first part and I hate leaving things halfway done).
Sometime in Q3 2014 (whatever that means) starts Harvard's Muscoloskeletal Anatomy − a good complement to Duke's Physiology, with virtual labs (including dissections). Also, if I see the Systems Biology through, then I'll probably embark on Experimental Methods in Systems Biology and the rest of the specialization.
So, that promises to be quite an eventful second half of the year. We'll see how it turns out!
On a side note, I'm starting to wonder about turning all this knowledge into an actual degree and a career change. There seems to be options with the CNAM (France's continuing education university-like institution, aimed at working professionals) so I'll be contacting them in September to study things through. Paradoxically enough, if I do end up taking evening classes there, it'll be because of MOOCs − but it'll also mean I can't take MOOCs anymore for time reasons. Oh well, that's all very hypothetical. We'll see!
Monday, July 21, 2014
Quantitative biology workshop, MIT
MIT's 7.QBW's winding down. I'm not sure what to think about it.
On the one hand, it's nice to get to grips with some biological problems with the tools actual computational biologists use. As usual, MIT do things seriously, the syllabus is impressive, the lectures are great and stimulating, etc.
On the other hand, it's very frustrating − we only get to lightly touch on to some very basic concepts. I can't really say I've learned anything of substance; most of the “workshops” have been, by necessity, hobbled by the idea that primarily biologists would take the course, with little to no competence in computer programming. And so, to review a number of languages and tools in a very short time, everything is kept very basic, very introductory. Only once so far was there a really interesting problem (the MATLAB one on neurological analysis).
A quick trawl through the forums show that basically there are two populations − the biologists who find the programming assignments very hard, and the programmers who breeze through the course and are frustrated not to do more biology.
I guess the course's ambition makes it a bit of a tightrope exercise; maybe the MOOC format isn't so well suited to that kind of course? I guess people would have found it more satisfying if the course had been either an applied programming crash course for biologists, or a biology application aimed at software people. Then, by excluding one half of the prospective population, they could have gone deeply enough to teach actual skills and knowledge to the other half.
On the third hand, by applying standard expectations, we are reading this wrong. It's not a “course” and it's not meant to teach skills'n'knowledge. It's a workshop, based on an outreach program. It's aimed at giving people a glimpse of what can be done in the field of computational biology − and in that respect, I'd say it's pretty well hit the nail on the head.
But then again, it's pretty frustrating to get a glimpse of something cool and not have any way to reach towards it. If MIT created a systems biology course or sequence of courses, then this workshop would be the best possible introduction to it. On its own, it's kinda… unfinished.
PS oh, yeah, of course, I'm going to ace the course. But for the reasons explained above, I have no real merit in it.
Update 2014-08-02: yay, here's my certificate:
(Yes, I shelled out for the Verified cert, because it goes some way towards contributing to edX and MIT, and mostly because there are rumours that MIT Biology are considering creating an XSeries − then have verified certs in likely courses is a good way to get a leg up.)
On the one hand, it's nice to get to grips with some biological problems with the tools actual computational biologists use. As usual, MIT do things seriously, the syllabus is impressive, the lectures are great and stimulating, etc.
On the other hand, it's very frustrating − we only get to lightly touch on to some very basic concepts. I can't really say I've learned anything of substance; most of the “workshops” have been, by necessity, hobbled by the idea that primarily biologists would take the course, with little to no competence in computer programming. And so, to review a number of languages and tools in a very short time, everything is kept very basic, very introductory. Only once so far was there a really interesting problem (the MATLAB one on neurological analysis).
A quick trawl through the forums show that basically there are two populations − the biologists who find the programming assignments very hard, and the programmers who breeze through the course and are frustrated not to do more biology.
I guess the course's ambition makes it a bit of a tightrope exercise; maybe the MOOC format isn't so well suited to that kind of course? I guess people would have found it more satisfying if the course had been either an applied programming crash course for biologists, or a biology application aimed at software people. Then, by excluding one half of the prospective population, they could have gone deeply enough to teach actual skills and knowledge to the other half.
On the third hand, by applying standard expectations, we are reading this wrong. It's not a “course” and it's not meant to teach skills'n'knowledge. It's a workshop, based on an outreach program. It's aimed at giving people a glimpse of what can be done in the field of computational biology − and in that respect, I'd say it's pretty well hit the nail on the head.
But then again, it's pretty frustrating to get a glimpse of something cool and not have any way to reach towards it. If MIT created a systems biology course or sequence of courses, then this workshop would be the best possible introduction to it. On its own, it's kinda… unfinished.
PS oh, yeah, of course, I'm going to ace the course. But for the reasons explained above, I have no real merit in it.
Update 2014-08-02: yay, here's my certificate:
Sunday, July 6, 2014
Summer is coming
Traditionally, summertime is break-time; students go home or travel to see the world or do summer jobs, that kind of thing. Traditions are thick-skinned and find themselves replicated in the world of MOOCs, where they don't have much sense (apart from letting staff take summer break too, I suppose).
So, let's round up:
Next week it's just these three MOOCs, then (just in time for my one-week holiday in Wales) I'm starting a Coursera MOOC on The Emergence of Life which sounds great. That's the whole of the summer pretty much accounted for; the serious stuff begins again in September, with Scottish Independence, Data Analysis and Inference, Physiology, Neural Data Analysis, Dino 101, Explore Statistics with R all starting together − and I'm reconsidering taking Systems Biology at Mount Sinai too (with 7.QBW under my belt, it should be easier). I won't be able to do all of them justice, and some will have to go: I think KI's R course is currently the most likely to be dropped, but it's not likely to be enough. I think either one of Data Analysis and Neural Data, or one of Physiology and Systems Biology, or both, will have to go. But we'll see; there's no harm (or indeed, cost) in trying all of them for a week or so, then picking the ones I like most.
So, let's round up:
- Introduction to Statistics is finished. I eventually got 89% in the first course, 65% in the second course (that I started on the one-but-last week and caught up), and 96% in the final course on Inference. That doesn't make a statistician of me, but I feel armed to use simple statistics to read the methods part of papers, for instance, which was approximately my ambition in starting the cycle. Overall, maybe the 15 weeks of the course could have been condensed into 10 or even fewer; and maybe doing one big course instead of three small ones would've been better. But I'm not really complaining.
- I'm down to three MOOCs now: Exoplanets, Genomic Medicine, and Quantitative Biology.
- Exoplanets (Astrophysics course number 2 from ANU) is great, I love the enthusiasm of the professors and the mysterious planet Mog.
- Genomic Medicine Gets Personal from Georgetown University is… strange. It's got loads of hand-holding, each video lecture is accompanied by reams of annoying text telling us what the pedagogical objectives are, etc.; the difficulty level is also very low and I can't really say I've been learning much actual science so far. But on the other hand, it's quite interesting to hear doctors explain case studies, and relate how technological progress really impacts their job.
- Quantitative Biology Workshop is great, and has rekindled my interest in systems biology. The course suffers from its format: it's nearly impossible to do more than simple introductory work in 6 weeks while covering as many different techniques and tools. But still: after this week's visual neuroscience exercise, I was quite happy to have achieved it, and starting to like MATLAB. I really hope that the hints dropped that MIT are considering a Biology Xseries turn out to be correct: so far their courses have been stellar.
Next week it's just these three MOOCs, then (just in time for my one-week holiday in Wales) I'm starting a Coursera MOOC on The Emergence of Life which sounds great. That's the whole of the summer pretty much accounted for; the serious stuff begins again in September, with Scottish Independence, Data Analysis and Inference, Physiology, Neural Data Analysis, Dino 101, Explore Statistics with R all starting together − and I'm reconsidering taking Systems Biology at Mount Sinai too (with 7.QBW under my belt, it should be easier). I won't be able to do all of them justice, and some will have to go: I think KI's R course is currently the most likely to be dropped, but it's not likely to be enough. I think either one of Data Analysis and Neural Data, or one of Physiology and Systems Biology, or both, will have to go. But we'll see; there's no harm (or indeed, cost) in trying all of them for a week or so, then picking the ones I like most.
Wednesday, July 2, 2014
Human sacrifice as a scientific technique
Actual test-yourself question from Astrophysics 2 - Exoplanets:
Which only proves the vital necessity for observatories of procuring a regular supply of fresh grad students.
That said, it's a nice illustration of why I like this course: there's serious stuff in there, but they don't forget their sense of humour. Another question asked whether the vibrations caused by nearby kangaroos jumping up and down might cause blurred spectral lines…
Subscribe to:
Posts (Atom)