#foundationsofappliedmath - Tagar Twitter

lebih dari 5 tahun yang lalu

Automatic (a.k.a. algorithmic) differentiation is an amazing tool. It can differentiate efficiently and accurately almost any function you can compute, including those defined by complicated algorithms #FoundationsOfAppliedMath https://t.co/fkAu560yKb

tj_jarvis's tweet photo. Automatic (a.k.a. algorithmic) differentiation is an amazing tool. It can differentiate efficiently and accurately almost any function you can compute, including those defined by complicated algorithms #FoundationsOfAppliedMath https://t.co/fkAu560yKb https://t.co/nZUeOuc8KE

1

8

0

3

0

Tyler Jarvis @tj_jarvis

lebih dari 5 tahun yang lalu

@m_letitbe @robertkaplinsky "Things you will use" and "train to think critically..." are not mutually exclusive. So why not spend our time on the things that do both? #FoundationsOfAppliedMath

1

5

0

Tyler Jarvis @tj_jarvis

lebih dari 5 tahun yang lalu

Newton's method converges in very few steps but each step is an expensive O(n^3) operation. BFGS reduces the cost of each step to O(n^2). It takes more steps to converge but in high dimensions can be faster overall #FoundationsOfAppliedMath @TheSIAMNews https://t.co/kmtdVLEb6C

tj_jarvis's tweet photo. Newton's method converges in very few steps but each step is an expensive O(n^3) operation. BFGS reduces the cost of each step to O(n^2). It takes more steps to converge but in high dimensions can be faster overall #FoundationsOfAppliedMath @TheSIAMNews https://t.co/kmtdVLEb6C https://t.co/8i7g7IjTol

0

12

1

2

0

Tyler Jarvis @tj_jarvis

lebih dari 5 tahun yang lalu

The degree-d polynomial interpolating at Chebyshev points in [-1,1] can be computed using the FFT in O(d log d) time, and the error shrinks rapidly with d, if the function being approximated is relatively smooth #FoundationsOfAppliedMath https://t.co/kmtdVLEb6C

tj_jarvis's tweet photo. The degree-d polynomial interpolating at Chebyshev points in [-1,1] can be computed using the FFT in O(d log d) time, and the error shrinks rapidly with d, if the function being approximated is relatively smooth #FoundationsOfAppliedMath https://t.co/kmtdVLEb6C https://t.co/ZnrqIzrG6e

0

5

0

1

0

Tyler Jarvis @tj_jarvis

lebih dari 5 tahun yang lalu

Perron's theorem: A matrix A with all positive entries has a simple, real eigenvalue λ equal to the spectral radius r(A) and it has a positive eigenvector. All the eigenvalues lie in the ball B(ε,λ–ε) for small enough ε>0 #FoundationsOfAppliedMath

tj_jarvis's tweet photo. Perron's theorem: A matrix A with all positive entries has a simple, real eigenvalue λ equal to the spectral radius r(A) and it has a positive eigenvector. All the eigenvalues lie in the ball B(ε,λ–ε) for small enough ε>0 #FoundationsOfAppliedMath https://t.co/WSIdVWZv7u

1

11

1

2

0

Tyler Jarvis @tj_jarvis

lebih dari 6 tahun yang lalu

I am very flattered that @jamestanton likes these books by Emily Evans, @JeffHumpherys, and me. James is a genius teacher and inspires much of what I do on the classroom, in my writing, and beyond #FoundationsOfAppliedMath

James Tanton

@jamestanton

lebih dari 6 tahun yang lalu

These volumes already are classics in my mind! Brilliant, beautiful, astoundingly comprehensive masterpieces. (You can make out the title details if you enlarge the photo.)

1

4

1

0

3

0

Tweet Teratas untuk #foundationsofappliedmath

Tagar Terakhir Terlihat di Sotwe

Tren untuk Anda

Pengguna Paling Populer