Journal of Algebraic Statistics

The Journal of Algebraic Statistics is an international electronic journal devoted to algebraic aspects of statistical theory, methodology and applications.  JAS seeks to publish a wide range of research and review papers that address one of the following:

  • Mathematical aspects of statistical models, where algebraic, geometric and combinatorial insights can be useful to study the behavior of statistical procedures;
  • Development of new statistical models and methods with interesting algebraic or geometric properties;
  • Novel applications of algebraic and geometric methods in statistics. 
 
The Journal of Algebraic Statisitcs is covered by the following indexing services:
  • Mathematical Reviews (MR) MathSciNet
  • CrossRef
  • Arastirmax Scientific Publication Index
  • Thomson Reuters ESCI and Web of Science

Aim and Scope:
 
Algebraic statistics focuses on mathematical aspects of statistical models, where algebraic, geometric and combinatorial insights can be useful to study behavior of statistical procedures. This approach has a long history in statistics and it can be traced back to Sir R.A. Fisher who used Abelian groups for experimental design and Karl Pearson who used polynomial algebra to study Gaussian mixture models. Since the turn of the century, the field has expanded and re-focused on applications of algebraic geometry, commutative algebra, and geometric combinatorics to the study of statistical models primarily for discrete data.

Recent advances in algebraic statistics have broadened the field beyond the 'traditional' algebraic statistics that focused on contingency tables and experimental design. The field is currently expanding in several directions. First, the study of Gaussian models have become an important part of algebraic statistics. Second, its computational aspects relying on numerical algebraic geometry are advancing to address statistical questions crucial for validity of statistical inference. Third, singular learning theory enables to study asymptotic statistics for models with hidden variables. Fourth, establishing identifiability of statistical models become an integral part of algebraic statistics. Finally, there is a strong focus on applications of these methods to phylogenetics, machine learning, biochemical reaction networks, social sciences, economics, and ecological inference. 
 
The Journal of Algebraic Statistics seeks to publish new research articles in the broad area of algebraic statistics, both theoretical advancements in the field - broadly defined - and its new applications.  From time to time, review papers on an emerging topic will also be considered appropriate. 

Vol 8, No 1 (2017)

Table of Contents

Articles - regular submission

Kaie Kubjas, Zvi Rosen
PDF
Shaowei Lin
PDF
Satoshi Aoki, Takayuki Hibi
PDF