A Boltzmann machine is a type of neural network that learns through the use of equations based on thermodynamics (basically, it 'agitates' the set of connections, then settles into a more stable state, a process of 'annealing'). This paper came up last week in an internal NRC series on quantum algorithms. In a nutshell, "propose a new machine-learning approach based on quantum Boltzmann distribution of a quantum Hamiltonian" ("the Hamiltonian of a system is an operator corresponding to the total energy of that system").
Today: 9 Total: 102 [Direct link] [Share]
