A
Alexandru Niculescu-Mizil
Researcher at Princeton University
Publications - 49
Citations - 9027
Alexandru Niculescu-Mizil is an academic researcher from Princeton University. The author has contributed to research in topics: Supervised learning & Bayesian network. The author has an hindex of 21, co-authored 49 publications receiving 7401 citations. Previous affiliations of Alexandru Niculescu-Mizil include Cornell University & NEC.
Papers
More filters
Proceedings ArticleDOI
An empirical comparison of supervised learning algorithms
TL;DR: A large-scale empirical comparison between ten supervised learning methods: SVMs, neural nets, logistic regression, naive bayes, memory-based learning, random forests, decision trees, bagged trees, boosted trees, and boosted stumps is presented.
Proceedings ArticleDOI
Model compression
TL;DR: This work presents a method for "compressing" large, complex ensembles into smaller, faster models, usually without significant loss in performance.
Proceedings ArticleDOI
Predicting good probabilities with supervised learning
TL;DR: The relationship between the predictions made by different learning algorithms and true posterior probabilities is examined, showing that maximum margin methods such as boosted trees and boosted stumps push probability mass away from 0 and 1 yielding a characteristic sigmoid shaped distortion in the predicted probabilities.
Proceedings ArticleDOI
Ensemble selection from libraries of models
TL;DR: A method for constructing ensembles from libraries of thousands of models using forward stepwise selection to be optimized to performance metric such as accuracy, cross entropy, mean precision, or ROC Area is presented.
Proceedings ArticleDOI
Topic-link LDA: joint models of topic and author community
TL;DR: A Bayesian hierarchical approach is developed that performs topic modeling and author community discovery in one unified framework and is demonstrated on two blog data sets in different domains and one research paper citation data from CiteSeer.