Dynamics of the maximum marginal likelihood hyperparameter estimation in image restoration: Gradient descent versus expectation and maximization algorithm
スポンサーリンク
概要
- 論文の詳細を見る
Dynamical properties of image restoration and hyperparameter estimation are investigated by means ofstatistical mechanics. We introduce an exactly solvable model for image restoration and derive differentialequations with respect to macroscopic quantities. From these equations, we evaluate relaxation processes of thesystem to the equilibrium state. Our statistical mechanical approach also enables us to investigate the hyperparameterestimation by means of maximization of the marginal likelihood by using gradient descent and theexpectation and maximization algorithm from the dynamical point of view.
- The American Physical Societyの論文
The American Physical Society | 論文
- First-principles calculations of the geometry and electronic structure of electron- and hole-doped C60 in the field-effect transistor structure
- Superconducting characteristics in electron-doped layered hafnium nitride : 15N isotope effect studies
- Unconventional Superconductivity in Electron-Doped Layered Li0.48(THF)yHfNCl
- Electrical-resistivity and low-temperature specific-heat measurements of single crystals of thiospinel CuV2S4
- Kaon B Parameter from Quenched Lattice QCD