DPPMMEstimation

Estimating parameters of determinantal point processes using an MM algorithm. Provides examples of the paper published in TMLR:
Takahiro Kawashima, Hideitsu Hino, "Minorization-Maximization for Learning Determinantal Point Processes," Transactions on Machine Learning Research, November 2023.

Working Directory

Please set your working directory at scripts/.

Recommended Environments

All the program are implemented in Julia.

Language	Recommended ver.
Julia	≥ 1.8.0

Library	Recommended ver.
Random
LinearAlgebra
SparseArrays
Plots	≥ 1.36.1
UnicodePlots	≥ 3.3.1
StatsBase	≥ 0.33.21
Distributions	≥ 0.25.76
DataFrames	≥ 1.4.2
DataFramesMeta	≥ 0.13.0
Query	≥ 1.0.0
JLD2	≥ 0.4.30
DeterminantalPointProcesses	≥ 0.2.2
MatrixEquations	≥ 2.2.2
CSV	≥ 0.10.7
MAT	≥ 0.10.3
StatsPlots	≥ 0.15.5

How to Setup

> git clone https://github.com/ISMHinoLab/DPPMMEstimation.git && cd DPPMMEstimation
> julia

(@v1.8) pkg> activate .
  Activating project at `/path/to/DPPMMEstimation`

(DPPMMEstimation) pkg> instantiate
# the mandatory packages will be installed
# you can check the environment by `pkg> status`

Codes for the Example

File	Description
scripts/exec_toydata.jl	Example on the toy data
scripts/exec_nottingham.jl	Example on the Nottingham dataset
scripts/exec_amazon.jl	Example on the Amazon Baby Registry Dataset
scripts/aggr_results.jl	Aggregate an experimental result into a DataFrame

References

• Kawashima, T. and Hino, H. "Minorization-Maximization for Learning Determinantal Point Processes," Transactions on Machine Learning Research, November 2023.
• Mariet, Z. and Sra, S. "Fixed-point Algorithms for Learning Determinantal Point Processes," ICML2015.
• Gillenwater, J. A., Kulesza, A., Fox, E. and Taskar, B. "Expectation-Maximization for Learning Determinantal Point Processes," NeurIPS2014.
• Nottingham Music Database: https://abc.sourceforge.net/NMD/
• jukedeck/nottingham-dataset: https://github.com/jukedeck/nottingham-dataset

License

This repository is released under the GNU GPLv3 license.