6 ms·

> One of the largest and most popular courses in the Stanford computer science department is CS229 — Machine Learning, taught by the artificial intelligence exp

> One of the largest and most popular courses in the Stanford computer science department is CS229 — Machine Learning, taught by the artificial intelligence expert and entrepreneur Andrew Ng. This course did not exist until 2003, when Professor Ng taught it for the first time with 68 students, and very little like it existed anywhere on college campuses 15 years ago. Today, the machine learning courses at Stanford enroll more than a thousand students. Machine Learning was a standard offering in CS departments in the early 2000's. It was already a venerable subfield with much of the basic theory having been developed in the 1970's and 1980's. I remember thinking it was something of a curiosity and not terribly practical but it's absurd to imply it sprung from nothing.

_delirium 7y ago
I was also surprised by that bit. Tom Mitchell's influential textbook [1] came out in 1997, and was already being used by quite a few courses within a few years of release. [1] http://www.cs.cmu.edu/~tom/mlbook.html http://www.cs.cmu.edu/~tom/mlbook.html
- noneeeed 7y ago
  That was my textbook at York Uni back in 2000, it was a great book. I recently got a new copy and quickly realised that I have forgotten most of the maths that I used to know, so first I need to do a refresher on that :) That book taught me an important lesson that I wish I'd learned earlier, that for many of my courses reading the textbook would have been a good idea. My lecturer for the ML course was unbelievably tedious, and I realised I'd absorbed almost nothing. I sat down with Mitchell's book for a week before the exam, did all the exercises, and got the highest mark in my degree. I often wonder how much better I could have done, and how much more I'd have learnt, if I'd put that sort of work into all my other modules earlier in the course.
danso 7y ago
How does your assertion contradict the quoted excerpt? It merely states that Ng's course began in 2003 with a relatively small number of students, and that few other colleges had taught such a course 15 years ago. Not that Ng invented the field of machine learning.
- laurencerowe 7y ago
  Specifically this part: > This course did not exist until 2003, when Professor Ng taught it for the first time with 68 students, and very little like it existed anywhere on college campuses 15 years ago. I took a Machine Learning course as part of my degree in 2001/2. About half of my year in the department attended. The course had been running for many years and I understood it to be a standard fixture in CS department curricula across the UK at the time.
  - danso 7y ago
    Thanks, that's all I was asking for (I won't demand that you do an empirical study across all colleges this early in the morning ;) )

_delirium 7y ago

I was also surprised by that bit. Tom Mitchell's influential textbook [1] came out in 1997, and was already being used by quite a few courses within a few years of release. [1] http://www.cs.cmu.edu/~tom/mlbook.html http://www.cs.cmu.edu/~tom/mlbook.html

noneeeed 7y ago
That was my textbook at York Uni back in 2000, it was a great book. I recently got a new copy and quickly realised that I have forgotten most of the maths that I used to know, so first I need to do a refresher on that :) That book taught me an important lesson that I wish I'd learned earlier, that for many of my courses reading the textbook would have been a good idea. My lecturer for the ML course was unbelievably tedious, and I realised I'd absorbed almost nothing. I sat down with Mitchell's book for a week before the exam, did all the exercises, and got the highest mark in my degree. I often wonder how much better I could have done, and how much more I'd have learnt, if I'd put that sort of work into all my other modules earlier in the course.

danso 7y ago

How does your assertion contradict the quoted excerpt? It merely states that Ng's course began in 2003 with a relatively small number of students, and that few other colleges had taught such a course 15 years ago. Not that Ng invented the field of machine learning.

laurencerowe 7y ago
Specifically this part: > This course did not exist until 2003, when Professor Ng taught it for the first time with 68 students, and very little like it existed anywhere on college campuses 15 years ago. I took a Machine Learning course as part of my degree in 2001/2. About half of my year in the department attended. The course had been running for many years and I understood it to be a standard fixture in CS department curricula across the UK at the time.
- danso 7y ago
  Thanks, that's all I was asking for (I won't demand that you do an empirical study across all colleges this early in the morning ;) )