Dynamic Determinantal Point Processes

Authors: Takayuki Osogami, Rudy Raymond, Akshay Goel, Tomoyuki Shirai, Takanori Maehara

AAAI 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Here, we apply the proposed learning algorithms to four music datasets (JSB Chorales, Nottingham, Piano-midi, and Muse Data), which have been used in Boulanger Lewandowski et al. (2012).
Researcher Affiliation Collaboration Takayuki Osogami, Rudy Raymond IBM Research AI Tomoyuki Shirai Institute of Mathematics for Industry Kyushu University Akshay Goel Graduate School of Mathematics Kyushu University Takanori Maehara RIKEN Center for Advanced Intelligence Project
Pseudocode Yes Algorithm 1 A learning algorithm for Dynamic DPPs. Algorithm 2 A learning algorithm for DPPs. Algorithm 3 Step 7 of Algorithm 1 with rank-one updates.
Open Source Code No The paper does not provide any explicit statements about making the source code available or include a link to a code repository.
Open Datasets Yes Here, we apply the proposed learning algorithms to four music datasets (JSB Chorales, Nottingham, Piano-midi, and Muse Data), which have been used in Boulanger Lewandowski et al. (2012).
Dataset Splits Yes The dataset is divided into training, validation, and test data.
Hardware Specification Yes We implemented our learning algorithms with Python and measured the computational time on a machine running Ubuntu 16.04 with 4.0 GHz Intel Core i7-6700K CPU and 48 GB Memory.
Software Dependencies No The paper mentions 'Python' but does not specify a version number for Python or any other software libraries or dependencies with their versions.
Experiment Setup Yes In Algorithm 1 and Algorithm 2, we choose the step size η via a relatively simple approach of backtracking line search, loosely following Armijo (1966)2. We stop the iteration when no step size significantly increases the objective function or when parameters are updated 100 times.