extending the black-scholes-merton constant volatility assumption
an exercise in itô processes and mathematical finance
anish lakkapragada
a sparse autoencoder is running live in your browser
▼
model trained on mnist with top-20 sparsity and running live with tf.js. inspired by this and that.
hidden layer units grid
top-4 activating inputs
feature —
supplement to the supplemental cs229 boosting lecture notes
testing a simple extension of adaboost to prevent overfitting
a quick tutorial of pac-bayesian and stability learning
the gentlest possible introduction to generalization error bounds beyond uniform convergence
using the vc dimension & rademacher complexity
evaluating some bounds to go from (statistical learning) theory to practice
exponential family discriminant analysis
generalizing linear discriminant analysis beyond normally distributed data
bernstein-von mises theorem & power posteriors
bayesian inference in model misspecification settings, visually explained
what if we didn't approximate posteriors?
a toy bayesian neural network with a precise parameter posterior
new website!
never escaping jekyll :)