Results 261 to 270 of about 3,624,996 (285)
Some of the next articles are maybe not open access.
2004
Assume we are trying to learn a concept class C of VC dimension d with respect to an arbitrary distribution. There is PAC sample size bound that holds for any algorithm that always predicts with some consistent concept in the class C (BEHW89): \(O(\frac{1}{\epsilon}(dlog\frac{1}{\epsilon}+log\frac{1}{\epsilon}))\), where e and δ are the accuracy and ...
openaire +1 more source
Assume we are trying to learn a concept class C of VC dimension d with respect to an arbitrary distribution. There is PAC sample size bound that holds for any algorithm that always predicts with some consistent concept in the class C (BEHW89): \(O(\frac{1}{\epsilon}(dlog\frac{1}{\epsilon}+log\frac{1}{\epsilon}))\), where e and δ are the accuracy and ...
openaire +1 more source
On the optimality of inclusion algorithms
1986In this paper a general concept of inclusion algorithm is introduced. Any inclusion algorithm provides a set that includes the solution of a given problem. Inclusion algorithms are studied with respect to the information used by them.
openaire +2 more sources
On the Optimality of Feautrier’s Scheduling Algorithm
Concurrency and Computation: Practice and Experience, 2002AbstractFeautrier's scheduling algorithm is the most powerful existing algorithm for parallelism detection and extraction, but it has always been known to be suboptimal. However, the question as to whether it may miss some parallelism because of its design has not been answered. We show that this is not the case.
openaire +3 more sources
1995
Most learning algorithms for single neuron are not able to provide for any classification problem the weight vector which satisfies the maximum number of input-output relations contained in the training set. An important exception is given by the pocket algorithm: it repeatedly executes the perceptron algorithm and maintains (in the pocket) the weight ...
openaire +4 more sources
Most learning algorithms for single neuron are not able to provide for any classification problem the weight vector which satisfies the maximum number of input-output relations contained in the training set. An important exception is given by the pocket algorithm: it repeatedly executes the perceptron algorithm and maintains (in the pocket) the weight ...
openaire +4 more sources
Battlefield Optimization Algorithm
Expert Systems with Applications, 2023Dadang Setiawan +3 more
openaire +1 more source
Mandrill optimization algorithm
Pattern Analysis and ApplicationsParsa Parsafar, Iman Esmaeili Afrokati
openaire +1 more source
Northern Goshawk Optimization: A New Swarm-Based Algorithm for Solving Optimization Problems
IEEE Access, 2021Pavel Trojovský +2 more
exaly
A Conceptual and Practical Comparison of PSO-Style Optimization Algorithms
Expert Systems With Applications, 2021Wolfram Schenck, Alaa Tharwat
exaly