Lausanne / Ho Chi Minh City
I build machines
that learn.
Machine Learning @ EPFL, working to make the world understand mathematics a little better — one gradient at a time.
click to teleport
f(x,y) = x² + 0.5y²
A perfect bowl. Gradient descent always converges to the global minimum at (0,0).
Loss
0.00e+0