580 California St., Suite 400
San Francisco, CA, 94104
Key research themes
1. How do stochastic optimal growth models extend classical deterministic growth theory, particularly through dynamic programming and continuous-time methods?
This theme focuses on the development and impact of stochastic optimal growth models as generalizations of classical deterministic growth theories such as Ramsey-Cass-Koopmans models. It investigates how stochastic formulations, both in continuous and discrete time, incorporate uncertainty and leverage dynamic programming tools to provide more realistic and flexible representations of economic growth processes. The theme highlights methodological innovations that enable deeper analyses of economic planning and business cycle dynamics under uncertainty.
Key finding: The paper traces the origins and impact of stochastic generalizations of classical growth models, particularly the Ramsey-Cass-Koopmans framework, highlighting how continuous-time and discrete-time stochastic optimal growth... Read more
2. What mathematical formalisms and unifying models best capture sigmoid growth phenomena across biological and economic systems?
Researchers have sought universal or generalized mathematical growth models capable of representing various sigmoid growth behaviors observed in biological populations, forestry, and economic systems. This theme examines the identification and application of flexible parametric formulas, including logistic, Gompertz, Bertalanffy, Richards, and their generalizations through Box-Cox transformations or q-statistics. It explores the development and comparative utility of such models for accurately fitting empirical data, their mathematical properties, and the implications for growth predictions and parameter interpretability.
Key finding: The paper presents a generalized sigmoid growth equation depending on two shape parameters that includes many known sigmoid models (e.g., Richards, logistic, Weibull) as special cases via transformations such as the Box-Cox.... Read more
Key finding: This study introduces a generalized growth model grounded in Tsallis nonextensive statistical mechanics, demonstrating that widely-used growth laws such as power law, exponential, logistic, Gompertz, Richards, and Von... Read more
Key finding: The work proposes novel criteria for selecting among sigmoid growth models (logistic, Gompertz, Bertalanffy) based on geometric measures: mean curvature and arc length of growth curves fitted to observed data. By computing... Read more
Key finding: The study presents a time-determinate asymmetric sigmoid growth model ('KM-function') that explicitly includes lifespan and enables flexible inflection points, differing from classical models implying infinite growth periods.... Read more
Key finding: This research systematically scans the parameter space of von Bertalanffy-Pütter type differential equations to identify optimal exponent pairs (a,b) that yield significantly better fits to individual mass-at-age growth data... Read more
3. How can growth in biological systems be realistically modeled incorporating randomness and individual variability using stochastic differential equations?
Recognizing intrinsic variability within biological populations, this research area applies stochastic differential equations (SDEs) and random logistic differential equations to capture individual-level randomness in growth rates and conditions. Techniques such as Bayesian inference to update random initial conditions, Karhunen-Loève expansions for stochastic process approximations, and solutions of Liouville's PDE for density evolution are employed. These methods enhance modeling of ecological, physiological, and population growth processes accounting for natural heterogeneity and uncertainty.
Key finding: This paper develops a biologically realistic stochastic growth model by extending the logistic differential equation to include random initial conditions and stochastic growth rates, capturing individual variability in... Read more
4. Which statistical and nonlinear regression approaches optimize growth curve fitting accuracy in livestock developmental studies?
Livestock growth modeling requires selecting appropriate nonlinear functions and statistical estimation methods to reliably predict key developmental parameters such as age at puberty and growth rates. This theme examines evaluation and comparison of multiple candidate models (e.g., Gompertz, logistic, von Bertalanffy, polynomial, exponential) using criteria like R², RMSE, AIC, and biological interpretability. Different parametrization techniques, including nonlinear least squares and Bayesian methods, are assessed for fit quality and predictive power in calves, cattle, and poultry.
Key finding: The study compares von Bertalanffy, logistic, and Gompertz models fitted to extensive birth-to-maturity weight data in Ongole Grade cattle. Results show all curves are sigmoid but differ in predicted puberty age and weight.... Read more
Key finding: Six nonlinear models were assessed for predicting preweaning weights in Holstein calves. All models fit data well, but the cubic polynomial function provided the best prediction performance across statistical criteria... Read more
Key finding: Re-analysing the same Holstein calf dataset as a prior study, this paper confirms that although several nonlinear growth functions can adequately predict preweaning weight, the cubic polynomial model consistently outperforms... Read more