11–20 of 161 posts
Is this a good start for Math required for Machine Learning ?
From what starting position are you asking from; what is your current math background?
Is this a good start for Math required for Machine Learning ?
No, this is for a general introduction to the mathematics of computer science. This looks like a basic (and good!) text any MIT freshman should be able to master. Perhaps it's for what 6.001 has morphed into? If you understand this stuff, you really need linear algebra for today's "deep learning", which perhaps is 18.03 (I can no longer remember).
18.06 is linear algebra. 18.03 is differential equations, which you don't really need for machine learning.
Earlier quoted context omitted.
From what starting position are you asking from; what is your current math background?
I am familiar with Linear algebra and at most average understanding of graph theory. I know there is lot more math to cover for ML but I find it overwhelming to start
Once you feel comfortable with those, you'll be more than ready to tackle 6.867: https://ocw.mit.edu/courses/electrical-engineering-and-compu....
Earlier quoted context omitted.
No, this is for a general introduction to the mathematics of computer science. This looks like a basic (and good!) text any MIT freshman should be able to master. Perhaps it's for what 6.001 has morphed into? If you understand this stuff, you really need linear algebra for today's "deep learning", which perhaps is 18.03 (I can no longer remember).
This class is 6.042 and is not required for CS majors. 6.001 morphed into 6.01, which is something to do with python. 18.06 is linear algebra. 18.03 is differential equations, which you don't really need for machine learning.
Is Gilbert Strang still teaching linear algebra at MIT? His intro materials to all things applied math are incredibly accessible.
Earlier quoted context omitted.
This class is 6.042 and is not required for CS majors. 6.001 morphed into 6.01, which is something to do with python. 18.06 is linear algebra. 18.03 is differential equations, which you don't really need for machine learning.
> 18.06 is linear algebra Is Gilbert Strang still teaching linear algebra at MIT? His intro materials to all things applied math are incredibly accessible.
Strang doesn't teach 18.06 every semester anyway.
Is this a good start for Math required for Machine Learning ?
Needs more: Linear Algebra: http://joshua.smcvt.edu/linearalgebra/ Calculus: https://cnx.org/contents/i4nRcikn@2.48:H2TLb2-S@2/Introducti... https://cnx.org/contents/HTmjSAcf@2.40:rrzms6rP@2/Introducti... https://cnx.org/contents/oxzXkyFi@2.49:72YaCFgv@2/Introducti...
Lucky I saw this!!
I do have one reservation though -- many of our students come in with a weaker mathematical background than MIT students; for example we spent several weeks doing proofs by induction (and no other kinds of proofs) and this text doesn't seem to feature a couple of weeks worth of examples.
I think I'll probably go with this and supplement as needed. Really it looks quite wonderful. (And hell, the book seems to be open source which would mean that I could potentially write supplmentary material directly into the book and make my version available publicly as well.)
This thread seems like a particularly good place to solicit advice: experiences with this book or others, what you wished you'd learned in your own undergraduate course on this subject, etc. I've taught this course once before -- I feel I did quite well but I still have room to improve. Thanks!
[1] https://www.amazon.com/Discrete-Mathematics-Applications-Sus...