The Duality between the Perceptron Algorithm and the von Neumann Algorithm
نویسندگان
چکیده
The perceptron and the von Neumann algorithms were developed to solve Linear Feasibility Problems. In this paper, we investigate and reveal the duality relationship between these two algorithms. The specific forms of Linear Feasibility Problems solved by the perceptron and the von Neumann algorithms are a pair of alternative systems by the Farkas Lemma. A solution of one problem serves as an infeasibility certificate of its alternative system. Further, we adapt an Approximate Farkas Lemma to interpret the meaning of an approximate solution from its alternative perspective. The Approximate Farkas Lemma also enables us to derive bounds for the distance to feasibility or infeasibility from approximate solutions of the alternative systems. Based on these observations, we interpret variants of the perceptron algorithm as variants of the von Neumann algorithm, and vice-versa; as well as transit the complexity results from one family to the other.
منابع مشابه
An Example with Decreasing Largest Inscribed Ball for Deterministic Rescaling Algorithms
Recently, Peña and Sohèili presented a deterministic rescaling perceptron algorithm and proved that it solves a feasible perceptron problem in O(mn log(ρ−1)) perceptron update steps, where ρ is the radius of the largest inscribed ball. The original stochastic rescaling perceptron algorithm of Dunagan and Vempala is based on systematic increase of ρ, while the proof of Peña and Sohèili is based ...
متن کاملCalculating Different Topological Indices of Von Neumann Regular Graph of Z_(p^α )
By the Von Neumann regular graph of R, we mean the graph that its vertices are all elements of R such that there is an edge between vertices x,y if and only if x+y is a von Neumann regular element of R, denoted by G_Vnr (R). For a commutative ring R with unity, x in R is called Von Neumann regular if there exists x in R such that a=a2 x. We denote the set of Von Neumann regular elements by V nr...
متن کاملAn Efficient Algorithm for Reducing the Duality Gap in a Special Class of the Knapsack Problem
A special class of the knapsack problem is called the separable nonlinear knapsack problem. This problem has received considerable attention recently because of its numerous applications. Dynamic programming is one of the basic approaches for solving this problem. Unfortunately, the size of state-pace will dramatically increase and cause the dimensionality problem. In this paper, an efficient a...
متن کاملBenson's algorithm for nonconvex multiobjective problems via nonsmooth Wolfe duality
In this paper, we propose an algorithm to obtain an approximation set of the (weakly) nondominated points of nonsmooth multiobjective optimization problems with equality and inequality constraints. We use an extension of the Wolfe duality to construct the separating hyperplane in Benson's outer algorithm for multiobjective programming problems with subdifferentiable functions. We also fo...
متن کاملThe James and von Neumann-Jordan type constants and uniform normal structure in Banach spaces
Recently, Takahashi has introduced the James and von Neumann-Jordan type constants. In this paper, we present some sufficient conditions for uniform normal structure and therefore the fixed point property of a Banach space in terms of the James and von Neumann-Jordan type constants and the Ptolemy constant. Our main results of the paper significantly generalize and improve many known results in...
متن کامل