Linear Hiearchical Models
Understanding generalized linear mixed models and MCMC sampling techniques.
- Part 1: Bayesian linear regression using Gibbs sampling
- Part 2: Regression with cluster effect in Python
- Part 3: Linear hiearchical models with spatial random effect
- Part 4: Linear hiearchical models with augmented spatial random effect
- Part 5: GLM with nested random effects
- Part 6: Poisson log-linear model
Deep Learning Fun
Various things involving neural networks.
- Part 1: Back propagation in a neural network
- Part 2: Feed-foward neural network using Numpy
- Part 3: Feed-foward neural network using TensorFlow
- Part 4: Convolutional neural network using TensorFlow
My three attempts to tackle a binary classification problem of chest X-ray images.
- Part 1: Data preparation, logistic regression
- Part 2: Feed-forward neural networks
- Part 3: Convolutional neural networks
A classification web app that gives predictions, cross-validation score, and principal component plot.
The user can upload a clean csv file and select the features and response variables. The app will fit a logistic regression through the data. More methods will be introduced in the future. Update: Now include both 2D and 3D PCA plots! 7/17/2018
Cognition Study of Multiple Myeloma Patients
A longitudinal study in multiple myeloma patients.
Assessment of cognition using cognitive domains of the National Institues of Health (NIH) Toolbox. Additionally, a quality of life measure, FACT-MM will be used.
Linear Transformation in 2D
A simple web app that illustrates the beauty of linear transformation in the plane.
The user can enter any 2x2 matrix and visualize how it transforms the unit circle and the eigenvectors.
Visualize Gamma Distribution
A simple web app that visualizes the shape of Gamma distribution.
By changing the parameters (alpha and beta), the user will learn how the shape of the Gamma distribution is affected.
Analysis of Lipid Levels Among College Students
Analysis of complex survey data from NHANES.
Using weighted analysis, we obtain point and interval estimates of lipid levels among college students.