TL;DR
This paper introduces a straightforward penalized likelihood method for estimating composition on discrete domains, addressing practical issues like smoothing and prior use, with implementation in R for practitioners.
Contribution
It presents a novel, simple penalized likelihood approach for composition estimation, including practical guidance and theoretical insights.
Findings
Effective smoothing parameter selection demonstrated
Incorporation of prior information improves estimates
Method implemented in accessible R functions
Abstract
In this note, we explore a simple approach to composition estimation, using penalized likelihood density estimation on a nominal discrete domain. Practical issues such as smoothing parameter selection and the use of prior information are investigated in simulations, and a theoretical analysis is attempted. The method has been implemented in a pair of R functions for use by practitioners.
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsStatistical Methods and Inference · Statistical Methods and Bayesian Inference · Bayesian Methods and Mixture Models