generic exception of cma module

A box in R^n. I.e., each coordinate is bounded.

Trust Region Policy Optimization

{0,1,...,n-1}

``CMAOptions()`` returns a dictionary with the available options and their default values for class ``CMAEvolutionStrategy``. ``CMAOptions('p

Vanilla Policy Gradient.

A utility class for experience replay. The code is adapted from https://github.com/spragunr/deep_q_rl

data logger for class `CMAEvolutionStrategy`. The logger is identified by its name prefix and (over-)writes or reads according data files. The

Performs constrained optimization via line search. The search direction is computed using a conjugate gradient algorithm, which gives x = A^{

Performs constrained optimization via penalized L-BFGS. The penalty term is adaptively adjusted to make sure that the constraint is satisfied

blanc container class for having a collection of attributes, that might/should at some point become a more tailored class

Performs unconstrained optimization via L-BFGS.

Performs unconstrained optimization via L-BFGS.

Trust Region Policy Optimization

Performs constrained optimization via line search. The search direction is computed using a conjugate gradient algorithm, which gives x = A^{

Performs (stochastic) gradient descent, possibly using fancier methods like adam etc.

This strategy implements the Ornstein-Uhlenbeck process, which adds time-correlated noise to the actions taken by the deterministic policy.

Performs constrained optimization via penalized L-BFGS. The penalty term is adaptively adjusted to make sure that the constraint is satisfied

container to keep track of the best solution seen

CMA-ES stochastic optimizer class with ask-and-tell interface. Calling Sequences ================= es = CMAEvolutionStrategy(x0, sig

Deep Deterministic Policy Gradient.

using ``time.clock`` with overflow handling to measure CPU time. Example: >>> clock = ElapsedTime() # clock starts here >>> t1 = clock(

Rotation class that implements an orthogonal linear transformation, one for each dimension. By default reach ``Rotation`` instance provides a

Trust Region Policy Optimization

keep and update a termination condition dictionary, which is "usually" empty and returned by `CMAEvolutionStrategy.stop()`. The class methods

lasagne.layers.BatchNormLayer(incoming, axes='auto', epsilon=1e-4, alpha=0.1, mode='low_mem', beta=lasagne.init.Constant(0), gamma=lasagn