Reputation
13,169
Top tag
Next privilege 15,000 Rep.
Protect questions
Badges
3 20 49
Newest
 Necromancer
Impact
~291k people reached

2h
comment Is the area of linear programming dead right now?
No, it's not dead. The development of algorithms for solving very large scale convex optimization problems (including linear programs) is an active area of research. For example, in recent years new algorithms for solving linear programs arising in compressed sensing have been developed. LPs are classically solved with interior point methods (or the simplex method), but for very large scale problems these classical methods are too expensive and so we might try using proximal algorithms, a current hot topic in convex optimization research.
7h
comment Prerequisites for Spivak's Calculus on Manifolds
You also need to know some linear algebra.
1d
comment A good companion to Axler's “Linear Algebra Done Right”?
Yes, this book certainly doesn't go overboard with applications. It does have a few nice applications, such as sections on special relativity, solving homogeneous linear systems of ODEs, and Markov chains.
2d
answered A good companion to Axler's “Linear Algebra Done Right”?
2d
comment Why do we teach Calculus in High School instead of, say, programming?
If physics and engineering aren't sufficient applications, machine learning and statistics make heavy use of calculus. And data science is one of the hottest jobs right now. However, in defense of this question, there is an article called The Calculus Trap on artofproblemsolving.com which points out that moving into calculus asap may not always be the best idea.
Feb
5
comment Why are math textbooks that are considered good books so hard to read? Why do authors make their books difficult to read?
One thing that may help to explain the style of baby Rudin and similar math books is to note that, while they may not be written in a way that you think is easy to learn from, they are written in a way that is very mathematically elegant. And mathematicians tend to place high value on this elegance, for aesthetic reasons. Check out other authors like Strang or Axler or Scott Aaronson, etc, for some styles you might find easier to learn from (while still being elegant).
Feb
3
comment what should I read after elements?
Tangential comment: it always strikes me as a bit strange that people still read Euclid as a textbook when we use modern textbooks for everything else. As great a genius as Euclid was, surely in the past 2000 years we've managed to write a better geometry textbook (building on Euclid's work, of course).
Feb
2
comment Linear Algebra book for beginners
I think you might have an easier time with Introduction to Linear Algebra, which is sometimes used as a textbook for lower division undergrad linear algebra courses.
Feb
2
awarded  Necromancer
Jan
26
comment Discrete Linear Programming over Finite Fields?
How are finite fields relevant to this? (Maybe they are, I'd just like to understand how. I've never heard of doing "linear programming over a finite field".)
Jan
25
comment How did Von Neumann think of the formula for scalar product?
This is a well known identity for the 2-norm -- for example, this identity is used to prove that a linear transformation that preserves the 2-norm also preserves inner products. You can prove the identity easily (for the 2-norm) by expanding the right hand side.
Jan
23
comment Book on applied mathematics
I think this book is written for applied math students just as much as it's written for engineering and physics students.
Jan
23
comment Calculus chain rule as “expanded form”
I'm actually not a fan of this way of stating the chain rule, because two different functions are both given the name $f$. Different functions should have different names.
Jan
23
answered Calculus chain rule as “expanded form”
Jan
23
answered Book on applied mathematics
Jan
22
comment Non-linear systems convergence
In convex optimization, convergence for many important iterative methods such as the Douglas-Rachford method is shown by expressing the method as fixed-point iteration for a certain nonlinear operator, then invoking convergence results for fixed-point iteration. One important convergence result is that if an operator is firmly nonexpansive and has a fixed point then fixed-point iteration converges to a fixed point. I'd be interested in learning what results are available for Gauss-Seidel iteration.
Jan
21
comment Why is $|x · y| ≤ ||x||_1||y||_∞$?
You could start by using the triangle inequality on the left hand side.
Jan
18
awarded  Famous Question
Jan
18
awarded  Necromancer
Jan
18
awarded  Nice Answer