Category Archives: Generalization Performance

LM101-077: How to Choose the Best Model using BIC

BIC model selection

Podcast: Play in new window | Download | Embed

Episode Summary: In this episode, we explain the proper semantic interpretation of the Bayesian Information Criterion (BIC) and emphasize how this semantic interpretation is fundamentally different from AIC (Akaike Information Criterion) model selection methods. Briefly, BIC is used to estimate the probability of the training data given the probability model, while AIC is used to estimate out-of-sample prediction… Read More »

LM101-076: How To Choose the Best Model using AIC or GAIC

The Model Selection Problem.

Podcast: Play in new window | Download | Embed

Episode Summary: In this episode, we explain the proper semantic interpretation of the Akaike Information Criterion (AIC) and the Generalized Akaike Information Criterion (GAIC) for the purpose of picking the best model for a given set of training data.  The precise semantic interpretation of these model selection criteria is provided, explicit assumptions are provided for the AIC and… Read More »

LM101-058: How to Identify Hallucinating Learning Machines using Specification Analysis

Robot imagines that he sees a pink elephant .

Podcast: Play in new window | Download | Embed

LM101-058: How to Identify Hallucinating Learning Machines using Specification Analysis Episode Summary: In this 58th episode of Learning Machines 101, I’ll be discussing an important new scientific breakthrough published just last week for the first time in the journal Econometrics  in the special issue on model misspecification titled “Generalized Information Matrix Tests for Detecting Model Misspecification”. The article… Read More »

LM101-055: How to Learn Statistical Regularities using MAP and Maximum Likelihood Estimation (Rerun)

Cartoon describing statistical learning with priors.

Podcast: Play in new window | Download | Embed

LM101-055: How to Learn Statistical Regularities using MAP and ML Estimation Episode Summary: In this rerun of Episode 10, we discuss fundamental principles of learning in statistical environments including the design of learning machines that can use prior knowledge to facilitate and guide the learning of statistical regularities. Show Notes: Hello everyone! Welcome to the tenth podcast in… Read More »

LM101-040: How to Build a Search Engine, Automatically Grade Essays, and Identify Synonyms using Latent Semantic Analysis

Mother and android looking at a park scene and learning about "dogs".

Podcast: Play in new window | Download | Embed

LM101-040: How to Build a Search Engine, Automatically Grade Essays, and Identify Synonyms using Latent Semantic Analysis Episode Summary: In this episode we explain how to build a search engine, automatically grade essays, and identify synonyms using Latent Semantic Analysis. Show Notes: Hello everyone! Welcome to the fortieth podcast in the podcast series Learning Machines 101. In this… Read More »

LM101-033: How to Use Linear Machine Learning Software to Make Predictions (Linear Regression Software)[RERUN]

Free Linear Regression Software Download

Podcast: Play in new window | Download | Embed

LM101-033: How to Use Linear Machine Learning Software to Make Predictions (Linear Regression Software)[RERUN]   Episode Summary: In this episode we describe how to download and use free linear machine learning software to make predictions for classifying flower species using a famous machine learning data set. This is a RERUN of Episode 13. Show Notes: Hello everyone! Welcome… Read More »

LM101-028: How to Evaluate the Ability to Generalize from Experience (Cross-Validation Methods)[RERUN]

There is no correct interpretation of the identity of an ambiguous letter.

Podcast: Play in new window | Download | Embed

LM101-028: How to Evaluate the Ability to Generalize from Experience (Cross-Validation Methods)[RERUN] Episode Summary: In this episode we discuss the problem of how to evaluate the ability of a learning machine to make generalizations and construct abstractions given the learning machine is provided a finite limited collection of experiences. Show Notes: Hello everyone! Welcome to a RERUN of… Read More »

LM101-027: How to Learn About Rare and Unseen Events (Smoothing Probabilistic Laws)[RERUN]

How do you compute the probability of rare and unseen events?

Podcast: Play in new window | Download | Embed

LM101-027: How to Learn About Rare and Unseen Events (Smoothing Probabilistic Laws)[RERUN] Episode Summary: In this podcast episode, we discuss the design of statistical learning machines which can make inferences about rare and unseen events using prior knowledge. Show Notes: Hello everyone! Welcome to a RERUN of the 11th podcast in the podcast series Learning Machines 101. In this… Read More »

LM101-013: How to Use Linear Machine Learning Software to Make Predictions (Linear Regression Software)

Android classifying a flower in the Iris Dataset.

Podcast: Play in new window | Download | Embed

Episode Summary: In this episode we describe how to download and use free linear machine learning software to make predictions for classifying flower species using a famous machine learning data set. Show Notes: Hello everyone! Welcome to the thirteenth podcast in the podcast series Learning Machines 101. In this series of podcasts my goal is to discuss important… Read More »

LM101-012: How to Evaluate the Ability to Generalize from Experience (Cross-Validation Methods)

There is no correct interpretation of the identity of an ambiguous letter.

Podcast: Play in new window | Download | Embed

Episode Summary: In this episode we discuss the problem of how to evaluate the ability of a learning machine to make generalizations and construct abstractions given the learning machine is provided a finite limited collection of experiences. Show Notes: Hello everyone! Welcome to the twelfth podcast in the podcast series Learning Machines 101. In this series of podcasts… Read More »