Detecting Anomalous Event Sequences with Temporal Point Processes
Authors: Oleksandr Shchur, Ali Caner Turkmen, Tim Januschowski, Jan Gasthaus, Stephan Günnemann
NeurIPS 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | In our experiments, we show that the proposed statistic excels at both traditional Go F testing, as well as at detecting anomalies in simulated and real-world data. |
| Researcher Affiliation | Collaboration | Oleksandr Shchur Technical University of Munich shchur@in.tum.de Ali Caner Türkmen Amazon Research atturkm@amazon.com Tim Januschowski Amazon Research tjnsch@amazon.com Jan Gasthaus Amazon Research gasthaus@amazon.com Stephan Günnemann Technical University of Munich guennemann@in.tum.de |
| Pseudocode | Yes | We provide the pseudocode description of our Oo D detection method in Appendix D. |
| Open Source Code | Yes | Code and datasets: https://github.com/shchur/tpp-anomaly-detection |
| Open Datasets | Yes | Code and datasets: https://github.com/shchur/tpp-anomaly-detection"; "LOGS: We generate server logs using Sock Shop microservices (Weave, 2017)"; "STEAD (Stanford Earthquake Dataset) (Mousavi et al., 2019) |
| Dataset Splits | Yes | We use the training set Dtrain to bit an RNN-based neural TPP model (Shchur et al., 2020) via maximum likelihood estimation (see Appendix F for details). Then, we de bone test statistics for the general TPP as follows."; "We split one large server log into 30-second subintervals, that are then partitioned into train and test sets. |
| Hardware Specification | Yes | The experiments were run on a machine with a 1080Ti GPU. |
| Software Dependencies | No | The paper mentions using an RNN-based neural TPP model and tools like Sock Shop microservices and Pumba, but does not provide specific version numbers for any software dependencies. |
| Experiment Setup | Yes | Details on the setup and datasets construction are provided in Appendix E & F."; "We use the training set Dtrain to bit an RNN-based neural TPP model (Shchur et al., 2020) via maximum likelihood estimation (see Appendix F for details). |