- Implemention :
KNN from Scratch; Stochastic Gradient Descent from Scracth; 3 layer Backpropogation Neural Network from Scracth;
- Application and Mathmatical Proof
-
Linear Regression
-
Guassian Discriminant Analysis/ Linear Discriminant Analysis
-
LASSO regression and Ridge regression
-
Tree based models( Random Forest/xgboost)
-
WordEmbedding - Word2Vec
regex expression /rmd report/ rmd latex
see the report here
It's a visualization project to visualize the time series data of australia weather and spatial map visualization from 2013 to 2017. publish at shinyaap.io changshen
1.The presentation slide available at here
including the database join(merge),cleaning, format standardization(from SAS to R), group and summarize
list some of the simple functions I write for statistical computing or visualization
Example
N <- c(100, 200, 300)
alpha <- .01
sd <- 5
delta <- seq(0.5, 5, 0.5)
mu1 <- 5
power.1<-Emperical.power(N, alpha, sd,delta,mu1)
View(power.1)
plot.emperical.power(power.1)
N <- c(20, 40)
alpha <- c(.05, .10)
sd <- c(0.5,1)
delta <- seq(0.1, 1, 0.1)
mu1 <- 2
power.2<-Emperical.power(N, alpha, sd,delta,mu1)
View(power.2)
plot.emperical.power(power.2)
a function for stratified randomization for a two-arm study Example: returned a list include stratified randomized data set, seed, distribution summary
testa <- stratified(t=2, s=4, samplesize = 100, equal = TRUE, seed = 89676);testa
testb <- stratified(t=3, s=3, samplesize = c(80, 70, 50), equal = FALSE, seed = 124589);testb
testc <- stratified(t=5, s=3, samplesize = 60, equal = TRUE, seed = 907563);testc