MASCOTTE no longer exists => visit the new project-team
Seminaire MASCOTTEQuantum-inspired Evolutionary Algorithms for Optimization
par C. Patvardhan (Dayalbagh Educational Institute Agra, INDIA)
| Date : | 15/06/10 | | Time : | 10:30 | | Location : | Galois Coriolis |
Quantum computing is the product of two of the most significant advances in the history of science: the theory of quantum mechanics, which describes the universe at the smallest scale with great accuracy, and the theory of computation, which has resulted in digital computers with exponentially increasing power and many efficient algorithms that are capable of solving a wide range of interesting problems. But, still, problems of slow/premature convergence remain and have to be tackled with suitable implementation for the particular problem at hand.<br />Quantum Evolutionary Algorithms (QEA) is a recent branch of EAs. QEA is a population-based probabilistic Evolutionary Algorithm that integrates concepts from quantum computing for higher representation power and robust search. It maintains a population of individuals in quantum bits or qubits. A qubit coded individual can probabilistically represent a linear superposition of states in the search space and has a better characteristic of population diversity than other representations.<br />Thus, QEAs are characterized by population dynamics, individual representation, evaluation function etc., as in EAs, as well as quantum bit (qubit) representation, superposition of states etc. as in Quantum Computing. The advantage of the QEAs is that, unlike the other EAs, they can work with small population sizes without being stuck in local minima and without converging prematurely because of loss of diversity. In the extreme case, the immense representation power of the qubits enables use of population size of 1. This reduces the computational burden and enables the solution of large sized problems.<br />The talk would introduce the QEAs and present some of our recent work on QEAs and applications.
Page des séminaires
| 
