pyepo.func.perturbed

Perturbed optimization function

Module Contents

Classes

perturbedOpt	An autograd module for Fenchel-Young loss using perturbation techniques. The
perturbedOptFunc	A autograd function for perturbed optimizer
perturbedFenchelYoung	An autograd module for Fenchel-Young loss using perturbation techniques. The
perturbedFenchelYoungFunc	A autograd function for Fenchel-Young loss using perturbation techniques.
implicitMLE	An autograd module for Implicit Maximum Likelihood Estimator, which yield
implicitMLEFunc	A autograd function for Implicit Maximum Likelihood Estimator
adaptiveImplicitMLE	An autograd module for Adaptive Implicit Maximum Likelihood Estimator, which
adaptiveImplicitMLEFunc	A autograd function for Adaptive Implicit Maximum Likelihood Estimator

Functions

_solve_or_cache(ptb_c, module)
_solve_in_pass(ptb_c, optmodel, processes, pool)	A function to solve optimization in the forward pass
_cache_in_pass(ptb_c, optmodel, solpool)	A function to use solution pool in the forward/backward pass
_solveWithObj4Par(perturbed_costs, args, model_type)	A global function to solve function in parallel processors

class pyepo.func.perturbed.perturbedOpt(optmodel, n_samples=10, sigma=1.0, processes=1, seed=135, solve_ratio=1, dataset=None)

Bases: pyepo.func.abcmodule.optModule

An autograd module for Fenchel-Young loss using perturbation techniques. The use of the loss improves the algorithmic by the specific expression of the gradients of the loss.

For the perturbed optimizer, the cost vector needs to be predicted from contextual data and are perturbed with Gaussian noise.

Thus, it allows us to design an algorithm based on stochastic gradient descent.

Reference: <https://papers.nips.cc/paper/2020/hash/6bb56208f672af0dd65451f869fedfd9-Abstract.html>

forward(pred_cost)

Forward pass

class pyepo.func.perturbed.perturbedOptFunc(*args, **kwargs)

Bases: torch.autograd.Function

A autograd function for perturbed optimizer

static forward(ctx, pred_cost, module)

Forward pass for perturbed

Parameters:

• pred_cost (torch.tensor) – a batch of predicted values of the cost

• module (optModule) – perturbedOpt module

Returns:

solution expectations with perturbation

Return type:

torch.tensor

static backward(ctx, grad_output)

Backward pass for perturbed

class pyepo.func.perturbed.perturbedFenchelYoung(optmodel, n_samples=10, sigma=1.0, processes=1, seed=135, solve_ratio=1, reduction='mean', dataset=None)

Bases: pyepo.func.abcmodule.optModule

An autograd module for Fenchel-Young loss using perturbation techniques. The use of the loss improves the algorithmic by the specific expression of the gradients of the loss.

For the perturbed optimizer, the cost vector need to be predicted from contextual data and are perturbed with Gaussian noise.

The Fenchel-Young loss allows to directly optimize a loss between the features and solutions with less computation. Thus, allows us to design an algorithm based on stochastic gradient descent.

Reference: <https://papers.nips.cc/paper/2020/hash/6bb56208f672af0dd65451f869fedfd9-Abstract.html>

forward(pred_cost, true_sol)

Forward pass

class pyepo.func.perturbed.perturbedFenchelYoungFunc(*args, **kwargs)

Bases: torch.autograd.Function

A autograd function for Fenchel-Young loss using perturbation techniques.

static forward(ctx, pred_cost, true_sol, module)

Forward pass for perturbed Fenchel-Young loss

Parameters:

• pred_cost (torch.tensor) – a batch of predicted values of the cost

• true_sol (torch.tensor) – a batch of true optimal solutions

• module (optModule) – perturbedFenchelYoung module

Returns:

solution expectations with perturbation

Return type:

torch.tensor

static backward(ctx, grad_output)

Backward pass for perturbed Fenchel-Young loss

class pyepo.func.perturbed.implicitMLE(optmodel, n_samples=10, sigma=1.0, lambd=10, distribution=sumGammaDistribution(kappa=5), two_sides=False, processes=1, solve_ratio=1, dataset=None)

Bases: pyepo.func.abcmodule.optModule

An autograd module for Implicit Maximum Likelihood Estimator, which yield an optimal solution in a constrained exponential family distribution via Perturb-and-MAP.

For I-MLE, it works as black-box combinatorial solvers, in which constraints are known and fixed, but the cost vector need to be predicted from contextual data.

The I-MLE approximate gradient of optimizer smoothly. Thus, allows us to design an algorithm based on stochastic gradient descent.

Reference: <https://proceedings.neurips.cc/paper_files/paper/2021/hash/7a430339c10c642c4b2251756fd1b484-Abstract.html>

forward(pred_cost)

Forward pass

class pyepo.func.perturbed.implicitMLEFunc(*args, **kwargs)

Bases: torch.autograd.Function

A autograd function for Implicit Maximum Likelihood Estimator

static forward(ctx, pred_cost, module)

Forward pass for IMLE

Parameters:

• pred_cost (torch.tensor) – a batch of predicted values of the cost

• module (optModule) – implicitMLE module

Returns:

predicted solutions

Return type:

torch.tensor

static backward(ctx, grad_output)

Backward pass for IMLE

class pyepo.func.perturbed.adaptiveImplicitMLE(optmodel, n_samples=10, sigma=1.0, distribution=sumGammaDistribution(kappa=5), two_sides=False, processes=1, solve_ratio=1, dataset=None)

Bases: pyepo.func.abcmodule.optModule

An autograd module for Adaptive Implicit Maximum Likelihood Estimator, which adaptively choose hyperparameter λ and yield an optimal solution in a constrained exponential family distribution via Perturb-and-MAP.

For AI-MLE, it works as black-box combinatorial solvers, in which constraints are known and fixed, but the cost vector need to be predicted from contextual data.

The AI-MLE approximate gradient of optimizer smoothly. Thus, allows us to design an algorithm based on stochastic gradient descent.

Reference: <https://ojs.aaai.org/index.php/AAAI/article/view/26103>

forward(pred_cost)

Forward pass

class pyepo.func.perturbed.adaptiveImplicitMLEFunc(*args, **kwargs)

Bases: implicitMLEFunc

A autograd function for Adaptive Implicit Maximum Likelihood Estimator

static backward(ctx, grad_output)

Backward pass for IMLE

pyepo.func.perturbed._solve_or_cache(ptb_c, module)

pyepo.func.perturbed._solve_in_pass(ptb_c, optmodel, processes, pool)

A function to solve optimization in the forward pass

pyepo.func.perturbed._cache_in_pass(ptb_c, optmodel, solpool)

A function to use solution pool in the forward/backward pass

pyepo.func.perturbed._solveWithObj4Par(perturbed_costs, args, model_type)

A global function to solve function in parallel processors

Parameters:

• perturbed_costs (np.ndarray) – costsof objective function with perturbation

• args (dict) – optModel args

• model_type (ABCMeta) – optModel class type

Returns:

optimal solution

Return type:

list