BIAS: Bayesian Inference At Scale in Julia

Version:0.1.1 Requires:julia releases 0.4.0 or later Date:April 05, 2016 Author:Adham Beyki (abeyki@deakin.edu.au) Web site:http://github.com/adham/BIAS.jl License:MIT

Purpose

BIAS is an open source implementation of several Bayesian parametric and non-parametric hierarchical mixture models in julia. The implemented models include:

1. Bayesian Mixture Model (BMM)
2. Latent Dirichlet Allocation (LDA)
3. Dirichlet Process Mixture Model (DPM)
4. Hierarchical Dielectric Mixture Model (HDP)
5. dynamic Hierarchical Dirichelt Mixture Model (dHDP)
6. Recurrent Chinese Restaurant Process (RCRP) also known as Temporal Dirichelt Process Mixture model (TDPM)

Getting Started

The following julia command will install the package:

julia> Pkg.clone("git://github.com/adham/BIAS.jl.git")

Contents

• Introduction
  • Mixture Models
• Distributions
  • Gaussian1D
  • Dirichlet
• Conjugates
  • Conjugate distributions
  • Interface
• Models
  • Bayesian Finite Mixture Model
  • Latent Dirichlet Allocation
  • Dirichlet Process Mixture Model
  • Hierarchical Dirichlet Process Mixture Model
  • dynamic Hierarchical Dirichlet Process Mixture Model
  • Temporal Dirichlet Process Mixture Model
• Examples
  • Bayesian Mixture Model
  • Latent Dirichlet Allocation
  • Recurrent Chinese Restaurant Process
• References

Indices

• Index