Everything about Discrete diffusion model
Technique for the generative modeling of a discrete probability distribution
In machine learning, discrete diffusion models are a class of diffusion models, which themselves are a class of latent variable generative models. Each discrete diffusion model consists of two major components: the forward jump diffusion process, and the reverse jump diffusion process. The goal of diffusion modeling is, given a given dataset and a forward process, to learn a model for the reverse process, such that the reverse process can generate new elements that are distributed similarly as the original dataset. A trained discrete diffusion model can be sampled in many ways, which trades off computational efficiency and sample quality. In general, higher quality data can be obtained, but at the price of higher computational cost. Wikipedia â
đ§ Start here: a 6-step reading path
- Read the overview: Discrete diffusion modelA 5-minute orientation
- Skim a recent paper: General Methods to Synchronize Fractional Discrete ReactionâDiffusion Systems Applied to the Glycolysis ModelInfluential work since 2022
- Pick up a classic book: Telling from discrete data whether the underlying continuous-time model is a diffusionThe most-reprinted book on the subject
- Read the landmark paper: Diffsound: Discrete Diffusion Model for Text-to-Sound GenerationCited 195 times
- Go sideways: Diffusion modelA closely related topic
- Go sideways: FokkerâPlanck equationA closely related topic
đ Definitions
No dictionary entry for this exact phrase.
đď¸ The subject at a glance
đ Landmark papers (most cited)
Diffsound: Discrete Diffusion Model for Text-to-Sound Generation
Viscous flow simulation using the discrete vortex modelâthe diffusion velocity method
On Stability of a Fractional Discrete ReactionâDiffusion Epidemic Model
A 2D discrete moisture diffusion model for simulating desiccation fracturing of soil
Pricing discrete barrier options under jump-diffusion model with liquidity risk
Existence of solutions for the discrete coagulation-fragmentation model with diffusion
Dissipative lattice model with exact traveling discrete kink-soliton solutions: Discrete breather generation and reaction diffusion regime
Mass-conserving solutions to the discrete coagulationâfragmentation model with diffusion
đ Influential since 2022
General Methods to Synchronize Fractional Discrete ReactionâDiffusion Systems Applied to the Glycolysis Model
Local Stability, Global Stability, and Simulations in a Fractional Discrete Glycolysis ReactionâDiffusion Model
G2P-DDM: Generating Sign Pose Sequence from Gloss Sequence with Discrete Diffusion Model
Diffusion and sedimentation in colloidal suspensions using multiparticle collision dynamics with a discrete particle model
Floor plan graph learning for generative design of residential buildings: a discrete denoising diffusion model
đ Books worth reading
- đTelling from discrete data whether the underlying continuous-time model is a diffusionYacine AĂŻt-Sahalia ¡ 2001
Ranked by number of editions on Open Library, a rough measure of how lasting a book is.
đ Related topics
- Diffusion model In machine learning, diffusion models, also known as diffusion-based generative models or score-based generative models, are a class of latent variableâŚ
- FokkerâPlanck equation In statistical mechanics and information theory, the FokkerâPlanck equation is a partial differential equation that describes the time evolution of theâŚ
- FeynmanâKac formula The FeynmanâKac formula, named after Richard Feynman and Mark Kac, establishes a link between parabolic partial differential equations and stochastic processesâŚ
- Diffusion process In probability theory and statistics, diffusion processes are a class of continuous-time Markov process with almost surely continuous sample paths. DiffusionâŚ
- McKeanâVlasov process In probability theory, a McKeanâVlasov process is a stochastic process described by a stochastic differential equation where the coefficients of the diffusionâŚ
- Kolmogorov backward equations (diffusion) The Kolmogorov backward equation (KBE) and its adjoint, the Kolmogorov forward equation, are partial differential equations (PDE) that arise in the theoryâŚ
- OnsagerâMachlup function The OnsagerâMachlup function is a function that summarizes the dynamics of a continuous stochastic process. It is used to define a probability densityâŚ
- Gaussian function In mathematics, a Gaussian function, often simply referred to as a Gaussian, is a function of the base form f(x)=expâĄ(âx2) and with parametric extensionâŚ