Covariate-Powered Empirical Bayes Estimation
Authors: Nikolaos Ignatiadis, Stefan Wager
NeurIPS 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We establish robust convergence guarantees for our method that hold under considerable generality, and exhibit promising empirical performance on both real and simulated data. |
| Researcher Affiliation | Academia | Nikolaos Ignatiadis Statistics Department Stanford University ignat@stanford.edu Stefan Wager Graduate School of Business Stanford University swager@stanford.edu |
| Pseudocode | No | The paper describes the algorithm steps in numbered text but does not present them as a formal pseudocode block or algorithm figure. |
| Open Source Code | Yes | The proposed EBCF (empirical Bayes with cross-fitting) method has been implemented in EBayes.jl (https://github.com/nignatiadis/EBayes.jl) |
| Open Datasets | Yes | The Movie Lens dataset consists of approximately 20 million ratings... [Harper and Konstan, 2016]... Communities and Crimes data from the UCI repository [Dua and Graff, 2017, Redmond and Baveja, 2002] |
| Dataset Splits | Yes | We randomly choose 10% of all users and attempt to estimate the movie ratings from them. This corresponds to having a much smaller dataset. We then summarize the i-th movie, by Zi, the average of the Ni users (in the training dataset) that rated it... XGBoost [Chen and Guestrin, 2016] with number of iterations chosen by 5-fold cross-validation |
| Hardware Specification | No | The paper does not provide specific details about the hardware (e.g., GPU/CPU models, memory) used for running the experiments. |
| Software Dependencies | No | The paper mentions 'EBayes.jl', 'MLJ.jl', 'Optim.jl', 'Distributions.jl', 'Julia language' and 'XGBoost', but does not consistently provide specific version numbers for all key software components within the text. |
| Experiment Setup | Yes | XGBoost [Chen and Guestrin, 2016] with number of iterations chosen by 5-fold cross-validation and η = 0.1 (weight with which new trees are added to the ensemble) |