Skip to content

Latest commit

ย 

History

14 Commits

Folders and files

NameName
Last commit message
Last commit date
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 

Repository files navigation

Imputing-Missing-Reaction-Times-Using-the-EM-Algorithm-Under-a-Bivariate-Normal-Model

This repository contains an R Markdown analysis where missing reaction time data for two stimuli (Stimulus A and Stimulus B) is imputed using the Expectation-Maximization (EM) algorithm under the assumption of a bivariate normal distribution.

๐Ÿ“ Project Overview

  • ๐Ÿ“„ Title: Imputing Missng Reaction Times Using the EM Algorithm
  • ๐Ÿงช Method: Expectation-Maximization (EM) Algorithm
  • ๐Ÿ“Š Data: Reaction time dataset from Duke University

๐Ÿ› ๏ธ Features

  • Custom EM algorithm implementation in R
  • Imputes values based on conditional expectations under multivariate normality
  • Tracks convergence through log-likelihood evaluation
  • Compares mean and variance before and after imputation
  • Preserves correlation structure between variables
  • Comprehensive visualizations including:
    • Histograms
    • Boxplots
    • Scatter plots (before and after imputation)

๐Ÿ“ฆ Dependencies

Make sure the following R libraries are installed:

library(ggplot2)
library(gridExtra)
library(dplyr)

##local dev
1. open README.md on your local 

About

Imputing missing reaction times using EM algorithm under a bivariate normal data

Topics

Resources

Stars

0 stars

Watchers

1 watching

Forks

Releases

Packages

Contributors