Non-conjugate Posterior using Stochastic Gradient Ascent with Adaptive Stepsize.
Kart-Leong LimPublished in: IJCNN (2020)
Keyphrases
- gradient ascent
- step size
- exponential family
- convergence rate
- cross entropy
- cost function
- expectation maximization
- bayesian framework
- neural network
- probability distribution
- posterior probability
- convergence speed
- probabilistic model
- particle swarm optimization
- markov decision processes
- markov chain monte carlo
- temporal difference
- dynamic programming
- reinforcement learning
- faster convergence
- multiscale