More Like this

1. Statistics of Mendelian segregation—A mixture model

   https://doi.org/10.1111/jbg.12394  

   Wang, Meiyue ; Xu, Shizhong ( April 2019 , Journal of Animal Breeding and Genetics)

   Mendel's law of segregation explains why genetic variation can be maintained over time. In diploid organisms, an offspring receives one allele from each parent, not just half of the blended genetic material of the parents. Which of the two alleles is received is purely random. This stochastic process generates genetic variation among members of the same family, called Mendelian segregation variance or within‐family variance. In statistics, the genetic value of a quantitative trait for an offspring follows a mixture distribution consisting of the four alleles of the two parents, guided by a Mendelian variable from each parent. The mixture model allows us to partition the total genetic variance into between‐family and within‐family variances. In the absence of inbreeding, the genetic variance splits half to the between‐family variance and half to the within‐family variance. With inbreeding, however, the between‐family variance is increased at the cost of the within‐family variance, leading to a net increase of the total genetic variance. This study defines multiple Mendelian variables and develops a mixture model of quantitative genetics. The phenomenon that allelic variance is maintained over time is guided by “the law of conservation of allelic variance” in biology, which is comparable to “the law of conservation of mass” in physics.

    

   more » « less
2. Fluidic Topology Optimization with an Anisotropic Mixture Model

   https://doi.org/10.1145/3550454.3555429  

   Li, Yifei ; Du, Tao ; Srinivasan, Sangeetha Grama ; Wu, Kui ; Zhu, Bo ; Sifakis, Eftychios ; Matusik, Wojciech ( December 2022 , ACM Transactions on Graphics)

   Fluidic devices are crucial components in many industrial applications involving fluid mechanics. Computational design of a high-performance fluidic system faces multifaceted challenges regarding its geometric representation and physical accuracy. We present a novel topology optimization method to design fluidic devices in a Stokes flow context. Our approach is featured by its capability in accommodating a broad spectrum of boundary conditions at the solid-fluid interface. Our key contribution is an anisotropic and differentiable constitutive model that unifies the representation of different phases and boundary conditions in a Stokes model, enabling a topology optimization method that can synthesize novel structures with accurate boundary conditions from a background grid discretization. We demonstrate the efficacy of our approach by conducting several fluidic system design tasks with over four million design parameters. 

   more » « less
3. Spectral clustering in the Gaussian mixture block model

   Li, Shuangping ; Schramm, Tselil ( April 2024 , arXivorg)
4. AdaMix: Mixture-of-Adaptations for Parameter-efficient Model Tuning

   https://doi.org/10.18653/v1/2022.emnlp-main.388  

   Wang, Yaqing ; Agarwal, Sahaj ; Mukherjee, Subhabrata ; Liu, Xiaodong ; Gao, Jing ; Awadallah, Ahmed Hassan ; Gao, Jianfeng ( January 2022 , Association for Computational Linguistics)
5. Understanding Symmetric Smoothing Filters: A Gaussian Mixture Model Perspective

   https://doi.org/10.1109/TIP.2017.2731208  

   Chan, Stanley H. ; Zickler, Todd ; Lu, Yue M. ( November 2017 , IEEE Transactions on Image Processing)