Valerio Gherardi's Personal Website

AIC in the well-specified linear model: theory and simulation

Model Selection
Linear Models
Regression
Statistics
R

Some illustrations of the Akaike Information Criterion (AIC) at work in a toy example.

"Annual adult survival rates for four sympatric breeding swallow species" by Imlay et al.

Comment on...
Population Dynamics
Biology
Statistics

An obscure mark-recapture data analysis.

Grammar as a biometric for Authorship Verification

Authorship Verification
Natural Language Processing
Forensic Science
Machine Learning
Statistics
R

Notes on preprint 2403.08462 by A. Nini, O. Halvani, L. Graner, S. Ishihara and myself.

"Induction and Deduction in Bayesian Data Analysis" by A. Gelman

Comment on...
Bayesian Methods
Statistics

On the importance of model checks in Bayesian data analysis.

"The Abuse of Power" by J. M. Hoenig and D. M. Heisey

Comment on...
Hypothesis Testing
Statistics

Why observed power calculations are useless (plus a few other points I don't buy).

AIC for the linear model: known vs. unknown variance

Model Selection
Linear Models
Regression
Statistics

Does knowledge of noise variance have any effect on model selection for the mean?

"A Closer Look at the Deviance" by T. Hastie

Comment on...
Maximum Likelihood Estimation
Linear Models
Statistics

A nice review of properties of Deviance for one parameter exponential families.

No binomial overdispersion from variations at the individual level

Population Dynamics
Biology
Ecology
Statistics

Some notes on the causes of overdispersion in count data.

On the first and second laws of thermodynamics for open systems

Open Systems
Thermodynamics
Physics

Matter transfer in open systems changes the relationship between heat and entropy, and work and volume.

Gravity waves in an ideal fluid

Atmospheric Physics
Fluid Dynamics
Waves
Physics

Compares the "parcel" method with standard linearization of fluid dynamics equations.

Binary digits of uniform random variables

Probability Theory

... are independent fair coin tosses.

Interpreting the Likelihood Ratio cost

Forensic Science
Bayesian Methods
Information Theory
Probability Theory
R

Analysis of infinite sample properties and comparison with cross-entropy loss.

Conditional Probability

Probability Theory
Measure Theory

Notes on the formal definition of conditional probability.

Prefix-free codes

Information Theory
Entropy
Probability Theory

Generalities about prefix-free (a.k.a. instantaneous) codes

AB tests and repeated checks

AB testing
Sequential Hypothesis Testing
Frequentist Methods
Statistics
R

False Positive Rates under repeated checks - a simulation study using R.

Testing functional specification in linear regression

Statistics
Model Misspecification
Regression
Linear Models
R

Some options in R, using the `{lmtest}` package.

Sum and ratio of independent random variables

Mathematics
Probability Theory

Sufficient conditions for independence of sum and ratio.

Fisher's Randomization Test

Statistics
Frequentist Methods
Causal Inference

Notes and proofs of basic theorems

p-values and measure theory

Probability Theory
Measure Theory
Frequentist Methods
Statistics

Self-reassurance that p-value properties don't depend on regularity assumptions on the test statistic.

Linear regression with autocorrelated noise

Statistics
Regression
Time Series
Linear Models
Model Misspecification
R

Effects of noise autocorrelation on linear regression. Explicit formulae and a simple simulation.

Model Misspecification and Linear Sandwiches

Statistics
Regression
Linear Models
Model Misspecification
R

Being wrong in the right way. With R excerpts.

Consistency and bias of OLS estimators

Statistics
Regression
Linear Models
Model Misspecification

OLS estimators are consistent but generally biased - here's an example.

Bayes, Neyman and the Magic Piggy Bank

Statistics
Confidence Intervals
Frequentist Methods
Bayesian Methods

Compares frequentist properties of credible intervals and confidence intervals in a gambling game involving a magic piggy bank.

Correlation Without Causation

Statistics

*Cum hoc ergo propter hoc*

How to get away with selection. Part II: Mathematical Framework

Statistics
Selective Inference
Model Misspecification

Mathematicals details on Selective Inference, model misspecification and coverage guarantees.

How to get away with selection. Part I: Introduction

Statistics
Selective Inference
R

Introducing the problem of Selective Inference, illustrated through a simple simulation in R.

kgrams v0.1.2 on CRAN

Natural Language Processing
R

kgrams: Classical k-gram Language Models in R.

R Client for R-universe APIs

R

{runi}, an R package to interact with R-universe repository APIs

Automatic resumes of your R-developer portfolio from your R-Universe

R

Create automatic resumes of your R packages using the R-Universe API.

{r2r} now on CRAN

Data Structures
R

Introducing {r2r}, an R implementation of hash tables.

Test post

Other

A short description of the post.

More articles »

Valerio Gherardi’s Personal Website

Corrections

If you see mistakes or want to suggest changes, please create an issue on the source repository.

Reuse

Text and figures are licensed under Creative Commons Attribution CC BY-SA 4.0. Source code is available at https://github.com/vgherard/vgherard.github.io/, unless otherwise noted. The figures that have been reused from other sources don't fall under this license and can be recognized by a note in their caption: "Figure from ...".