Recently Published
Meta analysis of prediction models
Meta analysis of diagnostic and prognostic models
Missing Data Handling in Meta Analyses
Adjustment for missing Data
Sensitivity Analysis Techniques
-Fixed equal
-Fixed opposite
-Random equal
-Random uncorrelated
Multiple Imputation Algorithm
Publication bias and small study effects handling in meta analyses
funnel plot
tests for small study effect
Modified tests for binary outcome
Adjusting for small study effect
Heterogeneity and meta regression
Measures of heterogeneity
Tests for subgroup differences
Meta regression with categorical/continuous covariate
Meta analyzing binary outcome
Binary effect measures: OR,RR,RD, Arc sine Difference
Estimating with sparse data points, Peto OR, FE model; inverse variance method, Mantel Haenzel method OR,RR,RD, Peto method, REmodel; DeSimonian-Laird method, Heterogeneity, subgroup analyses
Meta analysis core
Fixed effects and random effects model, modelling continuous outcomes, respective effect measures MD and SMD, Hartung-Knapp adjustment, prediction intervals, heterogeneity, subgroup analysis, meta-analyzing survival data, cross-over trials, and adjusted treatment effects are demonstrated using hypothetical/ simulated data.
Advanced feature slection in linear model with Boston housing data;best subsets, ridge,lasso regressions, elastic net and lasso with CV
A comparison of feature selection methods for the linear model including best subsets regression, ridge regression, lasso regression, elastic net, and lasso with cross validation.
Machine learning with MLR, LDA, QDA, ROC curve, AUC, cross validation and model selection
A data science and machine learning project using the Pima Indians Diabetes Data
machine learning model with logistic regression
This was created for a project on Coursera. Probably because the provided data had been simulated, a grossly overfitted model was resulted which fitted both train and test data with 100% accuracy. Despite the limitations, it provided leeway for experimenting with modelling including best-subsets selection in logistic regression models based on the AIC.