ProfessorArmando Teixeira-Pinto

Units of Study

• Machine Learning for Biostatistics - Master of Biostatistics
• Introductory Biostatistics
• Advanced Statistical Modelling
• Regression Modelling
• Introduction to Biostatistics in the module of Research Methods - Medical degree
• Controlled Clinical Trials

Teaching Resources

Brief tutorials

• Inverse probability weighting in R
• Linear Mixed Models in R
• Linear Mixed Models in jamovi
• Multiple Imputation in R

SIMULATIONS

• Neural Network

An animated, step-by-step backpropagation simulator built on a minimal two-neuron sigmoid network trained on one or two data points, where each of the six phases of a pass — two forward, loss, two backward, update — can be advanced individually with every formula shown with its numbers substituted, plus toggles for a second input, a second sample, SGD versus batch updating, and the learning rate.

• Randomisation

A simulator that recreates how randomisation tables were used in practice — pick a starting cell in a grid of random digits, apply a digit-to-allocation rule, and enrol patients one at a time under simple or blocked randomisation — then plots the running Treatment−Control imbalance and re-runs the same design 2,000 times to show the distribution of final imbalance, with teaching notes and discussion questions.

• Central Limit Theorem

An interactive Central Limit Theorem simulator in which students set a population distribution (normal, uniform, exponential, bimodal, or drawn by hand on the plot), repeatedly draw samples of size 5 to 100, and watch the sampling distribution of the mean or median accumulate against a fitted normal curve.

• Different MCMC algorithms

An animated MCMC dashboard that samples from a bivariate normal using Metropolis-Hastings, Gibbs, or Hamiltonian Monte Carlo, showing the chain's path in the joint space alongside the two marginal distributions as they build up, with adjustable number of draws, animation speed, and HMC leapfrog steps and step size.

• Linear regression with non-linear effects

An interactive demonstration of modelling a non-linear BMI–systolic blood pressure association within the linear model framework, letting students switch between polynomials, fractional polynomials, linear splines and restricted cubic splines, vary the degree, powers and number of knots, and compare the resulting fitted curves and R².

• Multivariable linear regression and correlation between covariates

A multiple linear regression simulator in which students set the sample size and the three pairwise correlations — between the two covariates, and between each covariate and the outcome — then inspect the fitted regression, the coefficient estimates and the residual plot as those correlations change.

• Multivariable linear regression with a binary and continuous covariates

An interactive demonstration that a multiple regression with one continuous and one binary covariate reduces in 2-D to two parallel fitted lines — one per level of X2 — with sliders for the X1–X2 correlation, the R², and the sample size.

• Coding dummy variables

An interactive demonstration on how dummy variables are coded from a categorical variable