• Journal of the Korean Society for Industrial and Applied Mathematics
• /
• v.28 no.1
• /
• pp.1-21
• /
• 2024

This review comprehensively explores the evolution, theoretical underpinnings, variations, and applications of diffusion models. Originating as a generative framework, diffusion models have rapidly ascended to the forefront of machine learning research, owing to their exceptional capability, stability, and versatility. We dissect the core principles driving diffusion processes, elucidating their mathematical foundations and the mechanisms by which they iteratively refine noise into structured data. We highlight pivotal advancements and the integration of auxiliary techniques that have significantly enhanced their efficiency and stability. Variants such as bridges that broaden the applicability of diffusion models to wider domains are introduced. We put special emphasis on the ability of diffusion models as a crucial foundation model, with modalities ranging from image, 3D assets, and video. The role of diffusion models as a general foundation model leads to its versatility in many of the downstream tasks such as solving inverse problems and image editing. Through this review, we aim to provide a thorough and accessible compendium for both newcomers and seasoned researchers in the field.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.10 no.16
• /
• pp.149-158
• /
• 1987

The Adoption/Diffusion of Innovations(New Product), a topic of study and research that has frown rapidly in the past few decades, deals with how a new product is adopted in a society. It is of high importance to marketing organizations because New Products must be brought out continuously in order to service. The purpose of this paper is to examine the Adoption/Diffusion Models for New product which will help to analyze the Adoption/Diffusion process of Adopters. There are a number of models that, with varying degrees of success, have been used to predict market acceptance of new product. In this paper, following types of new product Adoption/Diffusion Models was suggested. (1) Adoption Models : The Alternative Models of Adoption. The Rogers Model of the Innovation Decision Process. (2) Diffusion Models : First Purchase Models(Basic Models, Extension of the Basic Models), Repeat Purchase Models

• Journal of Information Technology Applications and Management
• /
• v.21 no.4_spc
• /
• pp.403-414
• /
• 2014

Forecasting technology in economic activity is a quite intricate procedure so researchers should grasp the point of the data to use. Diffusion models have been widely used for forecasting market demand and measuring the degree of technology diffusion. However, there is a question that a model, explaining a certain market with goodness of fit, always shows good performance with markets of different conditions. The primary aim of this paper is to explore diffusion models which are frequently used by researchers, and to help readers better understanding on those models. In this study, Logistic, Gompertz and Bass models are used for forecasting Global Wireline Subscribers and the performance of models is measured by Mean Absolute Percentage Error. Logistic model shows better MAPE than the other two. A possible extension of this study may verify which model reflects characteristics of industry better.

• Journal of the Korean Operations Research and Management Science Society
• /
• v.37 no.2
• /
• pp.113-125
• /
• 2012

The logistic model and the Bass model have diverse names and formulae in diffusion theory. This diversity makes users or readers confused while it also contributes to the flexibility of modeling. The method of handling the integration constant, which is generated in process of deriving the closed form solution of the differential equation for a diffusion model, results in two different 'actual' models. We rename the actual four models and propose the usage of the models with respect to the purpose of model applications. The application purpose would be the explanation of historical diffusion pattern or the forecasting of future demand. Empirical validation with 86 historical diffusion data shows that misuse of the models can draw improper conclusions for the explanation of historical diffusion pattern.

• Journal for History of Mathematics
• /
• v.29 no.6
• /
• pp.353-363
• /
• 2016

In mathematical population ecology which is an academic field that studies how populations of biological species change as times flows at specific locations in their habitats, PDE models have been studied in many aspects and found to have different properties from the classical ODE models. And different approaches to PDE type models in mathematical biology are still being tried currently. This article investigate various forms to express diffusion effects and review the history of PDE models involving diffusion terms in mathematical ecology. Semi-linear systems representing the spatial movements of each individual as random simple diffusion and quasi-linear systems describing more complex diffusions reflecting interspecific interactions are studied. Also it introduce a few of important problems to be solved in this field.

• Journal of the Korean Operations Research and Management Science Society
• /
• v.38 no.1
• /
• pp.201-216
• /
• 2013

Diffusion is a basic mathematical tool for modern financial engineering. The theory of the estimation methods for diffusion models is an important topic of the financial engineering. Many researches have been tried to apply the likelihood estimation method for estimating diffusion models. However, the likelihood estimation method for diffusion is complicated and needs much amount of computing. In this paper we develop the estimation methods which are simple enough to be compared to the Euler approximation method, and efficient enough statistically to be compared to the likelihood estimation method. We devise pseudo-likelihood and propose the maximum pseudo-likelihood estimation methods. The pseudo-likelihoods are obtained by approximating the transition density with normal distributions. The means and the variances of the distributions are obtained from the delta expansion suggested by Lee, Song and Lee (2012). We compare the newly suggested estimators with other existing estimators by simulation study. From the simulation study we find the maximum pseudo-likelihood estimator has very similar properties with the maximum likelihood estimator. Also the maximum pseudo-likelihood estimator is easy to apply to general diffusion models, and can be obtained by simple numerical steps.

• The Korean Journal of Applied Statistics
• /
• v.28 no.5
• /
• pp.895-906
• /
• 2015

Diffusion is a mathematical tool to explain the fluctuation of financial assets and the movement of particles in a micro time scale. There are ongoing statistical trials to develop an estimation method for diffusion models based on likelihood. When we estimate diffusion models by applying the maximum likelihood estimation method on data observed at discrete time points, we need to know the transition density of the diffusion. In order to approximate the transition densities of diffusion models, we suggests the method to approximate the path integral of the random process with normal random variables, and compare the numerical properties of the method with other approximation methods.

• Journal of Korean Institute of Industrial Engineers
• /
• v.14 no.1
• /
• pp.103-117
• /
• 1988

Replacement demand plays an important role to forecast the total demand of durable goods, while most of the diffusion models deal with only adoption data, namely initial purchase demand. This paper presents replacement demand forecasting models incorporating repurchase rate, multi-ownership, and dynamic product life to complement the existing diffusion models. The performance of replacement demand forecasting models are analyzed and practical guidelines for the application of the models are suggested when life distribution data or adoption data are not available.

• Journal of Hydrogen and New Energy
• /
• v.18 no.3
• /
• pp.265-274
• /
• 2007

A study has been conducted to clarify NO emission behavior through preferential diffusion effects of $H_2$ and H in methane-hydrogen diffusion flames. A comparison is made by employing three species diffusion models. Special concerns are focused on what is the deterministic role of the preferential diffusion effects in flame structure and NO emission. The behavior of maximum flame temperatures with three species diffusion models is not explained by scalar dissipation rate but the nature of chemical kinetics. The preferential diffusion of H into reaction zone suppresses the populations of the chain carrier radicals and then flame temperature while that of $H_2$ produces the increase of flame temperature. These preferential diffusion effects of $H_2$ and H are also discussed about NO emissions through the three species diffusion models.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.31 no.12
• /
• pp.1009-1016
• /
• 2007

Numerical study on preferential diffusion effects in flame structure in $CH_4-H_2$ diffusion flames is conducted with detailed chemistry. Comparison of flame structures with mixture-averaged species diffusion and suppression of the diffusivities of $H_2$ and H was made. Discernible differences in flame structures are displayed with three species diffusion models. The behaviors of maximum flame temperatures with those species diffusion models are not explained by scalar dissipation rate but by the nature of chemical kinetics. It is seen that the modifcation of flame structure is mainly due to the preferential diffusion of H2 and thereby the nature of chemical kinetics. It is also found that the behaviors of major species with the three species diffusion models are addressed to the nature of chemical kinetics, and this is evident by examining importantly contributing reaction steps to the production and destruction of those chemical species.