JSM 2022 후기

JSM 2022 후기

JSM은 Joint Statistical Meetings 의 약자로 전세계 통계학자들이 모이는 가장 큰 학회이다. 나는 이번 학회에 포스터 발표를 하게 되었는데, 다녀와서 느낀 점을 남겨보고자 한다.


- 8/7 (Sun)

 - Bayesian Modeling Of Spatial Transcriptomics via Gaussian Process (Boost-GP) : https://github.com/Minzhe/BOOST-GP

 - Ising model (Identify an SV gene via fitting an Ising model to dichotomized expression levels) : A model of interacting binary spins on a lattice.


- 8/8 (Mon)

 - A spatial Bayesian Latent Factor Model for Image-on-Image Regression by Jian Kang

  - Use rfMRI, sMRI to predict fMRI

  - Deep-learning: Image-to-image translation. Conditional adversarial network

  - Outcome image -> basis coefficient by Gaussian process -> latent variable by spatially latent model -> predicting image


- 8/9 (Tue)

 - In remembrance of the life and legacy of Marshall Joffe

 - Mortal Multi-Armed Bandit problem

 - Hangover ...

- 8/10 (Wed)

 - Faster Estimation for Constrained Gamma Mixture Models Using Close-Form Estimators by Jiangmei Xiong

  - Impose the location of the mode to derive the closed-form estimator for gamma distribution

- 8/11 (Thu)

 - Bayesian Spatial Binary Regression for Label Fusion in Structural Neuroimaging by Andrew Brown

  - Gaussian Markov Random Field to smooth out

- An All-In-One Statistical Framework That Generates Realistic Single-Cell Omics Data and Infers Cell Heterogeneity Structure by Jingyi Li (and my friend Dong Yuan Song)

 - scDesign3: transcriptomics 같은것도 simulation 가능해졌다 

(초기버젼: https://github.com/Vivianstats/scDesign)

- 느낀점

 - 다음 포스터 세션할 때는 컴퓨터를 가져와서 모델이 어떻게 작동하는지 보여주기 (Cell tracking method 특히)


- Museum of American History

- Museum of Arts

- Apple store in Carnegie library

- Rasika

- Capitol Hill

- Korean War Memorial

- Lab dinner at Zaytinya