Notice: This website is continuously updated as new conference proceedings and journal volumes are published. Consequently, the figures presented here may differ from those reported in our publications.
A brief overview of our methodology is provided on the Methods page, along with references to publications documenting our approach and links to the code and data used in each study.
The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates.
Scalable DBSCAN with Random Projections
Authors: HaoChuan Xu, Ninh Pham
NeurIPS 2024 | Venue PDF | LLM Run Details | Input Tokens: 34,501 Total number of tokens sent to the LLM as input for this paper's analysis. | Output Tokens: 5,927 Total number of tokens produced by the LLM (including reasoning/thinking tokens) for this paper's analysis.
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Empirically, s DBSCAN is significantly faster and provides higher accuracy than competitive DBSCAN variants on real-world million-point data sets. |
| Researcher Affiliation | Academia | Hao Chuan Xu School of Computer Science University of Auckland EMAIL Ninh Pham School of Computer Science University of Auckland EMAIL |
| Pseudocode | Yes | Algorithm 1 DBSCAN |
| Open Source Code | Yes | Our code is available at https://github.com/Ninh Pham/s Dbscan. |
| Open Datasets | Yes | We conduct experiments on three popular data sets: Mnist (n = 70, 000, d = 784, # clusters = 10), Pamap2 (n = 1, 770, 131, d = 51, # clusters = 18), and Mnist8m (n = 8, 100, 000, d = 784, # clusters = 10). |
| Dataset Splits | No | We conduct experiments on three popular data sets: Mnist (n = 70, 000, d = 784, # clusters = 10), Pamap2 (n = 1, 770, 131, d = 51, # clusters = 18), and Mnist8m (n = 8, 100, 000, d = 784, # clusters = 10). |
| Hardware Specification | Yes | We conducted experiments on Ubuntu 20.04.4 with an AMD Ryzen Threadripper 3970X 2.2GHz 32-core processor (64 threads) with 128GB of DRAM. |
| Software Dependencies | Yes | We implement s DBSCAN and s OPTICS in C++ and compile with g++ -O3 -std=c++17 -fopenmp -march=native. We use the Eigen library 2 for SIMD vectorization on computing the distances. |
| Experiment Setup | Yes | Parameter settings. We consider min Pts = 50 for all experiments. s DBSCAN and s OPTICS use D = 1024, m = min Pts. Randomized kernel embeddings use σ = 2ε, d = 1024. We use k = 5 for Mnist and k = 10 for Pamap2 and Mnist8m. |