PubMed Health⌕ Search
Home/Archive/PubMed

PubMed · 11052413

A discrete-time Lagrangian network for solving constrained quadratic programs.

W S Tang·J Wang
View original source ↗Export metadata ↓

Abstract

A discrete-time recurrent neural network which is called the discrete-time Lagrangian network is proposed in this letter for solving convex quadratic programs. It is developed based on the classical Lagrange optimization method and solves quadratic programs without using any penalty parameter. The condition for the neural network to globally converge to the optimal solution of the quadratic program is given. Simulation results are presented to illustrate its performance.

Explore related subjects

MathematicsNeural Networks, ComputerSoftwareLetter

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

W S Tang, J Wang. 2000. A discrete-time Lagrangian network for solving constrained quadratic programs.. https://doi.org/10.1142/s0129065700000260

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related citations

PubMed2002

Viscosity of concentrated suspensions: influence of cluster formation.

Dispersed particles can form clusters even at low concentrations. Colloidal and hydrodynamic forces are responsible for this phenomenon and these forces determine both structure and size of clusters. We assume that the viscosity of a concentrated suspension is completely determined by cluster size distribution, regardless if clusters form under the action of colloidal, hydrodynamic interactions or applied shear rates. Based on this assumption an equation, which describes dependency of viscosity on a concentration of dispersed particles taking into account cluster formation, is deduced. Under special restrictions the deduced dependency coincides with the well-known Dougherty-Krieger's equation except for a clear physical meaning of parameters entered. Our consideration shows that Dougherty-Krieger's equation has deeper physical background than it has been supposed earlier. Experimental verification of the suggested model shows a good agreement with the theory predictions and proves a presence of clusters even at low concentrations of dispersed particles.

Mathematics↗
PubMed2002

PK-PD curve-fitting problems with the Hill equation? Try one of the 1-exp functions derived from Hodgkin, Douglas or Gompertz.

UNLABELLED: Non-linear phenomena are observed with enzyme kinetics, protein binding, pharmacokinetics or pharmacodynamics. The Hill equation, the Michaelis-Menten equation extended by a power coefficient, is traditionally used for sigmoid curve fitting. Sigmoid saturation phenomena can also be described by exponential functions (1-exp), extended by a power coefficient such as those derived by Hodgkin, Douglas or Gompertz. Comparing the 4 equations, the sigmoid 1-exp function in the form of Hodgkin and Huxley comes closest to the principle of simplicity and succinctness with regard to definition, slope and flexibility of the inflection point. To compare the applicability, a standardized sample of 250 curves was generated by each I of the 4 equations and mutually fitted with the remaining 3. The Hill equation gives the closest fit with the data generated by the other functions. The Douglas variant exhibits the highest rate of convergence. The Gompertz function provides the basic feature of a baseline effect. CONCLUSION: The sigmoid functions investigated (Hill, Hodgkin, Douglas, Gompertz) have differing characteristics and can be used interchangeably for solving specific problems in non-linear modeling.

Mathematics↗