We propose a new class of rigorous methods for derivative-free optimization with the aim delivering efficient and robust numerical performance functions all types, from smooth to non-smooth, under different noise regimes. To this end, we have developed methods, called Full-Low Evaluation organized around two main types iterations. The first iteration type (called Full-Eval) is expensive in func...

Systematic backtracking is used in many constraint solvers and combinatorial optimisation algorithms. It is complete and can be combined with powerful search pruning techniques such as branchand-bound, constraint propagation and dynamic variable ordering. However, it often scales poorly to large problems. Local search is incomplete, and has the additional drawback that it cannot exploit pruning...

The conventional two-stage training algorithm of the fuzzy/neural architecture called FALCON may not provide accurate results for certain type of problems, due to the implicit assumption of independence that this training makes about parameters of the underlying fuzzy inference system. In this correspondence, a training scheme is proposed for this fuzzy/neural architecture, which is based on li...

We present the numerical solutions for the PDE-constrained optimization problem arising in cardiac electrophysiology, that is, the optimal control problem of monodomain model. The optimal control problem of monodomain model is a nonlinear optimization problem that is constrained by the monodomain model. The monodomain model consists of a parabolic partial differential equation coupled to a syst...

We present a novel definition of the reinforcement learning state, actions and reward function that allows a deep Q-network (DQN) to learn to control an optimization hyperparameter. Using Q-learning with experience replay, we train two DQNs to accept a state representation of an objective function as input and output the expected discounted return of rewards, or q-values, connected to the actio...

Criteria containing a barrier function i.e., an unbounded function at the boundary of the feasible solution domain are frequently encountered in the optimization framework. When an iterative descent method is used, a search along the line supported by the descent direction through the minimization of the underlying scalar function has to be performed at each iteration. Usual line search strateg...

Multi-point constructive search (MPCS) performs a series of resource-limited backtracking searches where each search begins either from an empty solution (as in randomized restart) or from a solution that has been encountered during the search. We perform a systematic study of MPCS to evaluate the performance impact of various parameter settings. Results using job shop scheduling instances with...

The steepest descent method proposed by Fliege and Svaiter has motivated the research on methods for multiobjective optimization, which received increasing attention in recent years. However, empirical results show that Armijo line search often a very small stepsize along direction, decelerates convergence seriously. This paper points out issue is mainly due to imbalances among objective functi...

We develop an optimization algorithm which is able to deal with inexact evaluations of the objective function. The proposed algorithm employs sequential quadratic programming with a line search that uses the l1 penalty function for an Armijo-like condition. Both the objective gradient computations for the quadratic subproblems and the objective function computations for the line search admit so...

Step size determination (also known as line search) is an important component in effective algorithmic development for solving the traffic assignment problem. In this paper, we explore a novel step scheme, Barzilai-Borwein (BB) size, and adapt it stochastic user equilibrium (SUE) The BB special scheme incorporated into gradient method to enhance its computational efficiency. It motivated by New...