Regional Concept Drift Detection and Density Synchronized Drift Adaptation
Authors: Anjin Liu, Yiliao Song, Guangquan Zhang, Jie Lu
IJCAI 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experimental evaluations on three benchmark data sets show that our concept drift adaptation algorithm improves accuracy compared to other methods. |
| Researcher Affiliation | Academia | Anjin Liu, Yiliao Song, Guangquan Zhang, Jie Lu Centre for Artificial Intelligence, School of Software Faculty of Engineering and Information Technology, University of Technology, Sydney, Australia {Anjin.Liu, Yiliao.Song}@student.uts.edu.au, {Guangquan.Zhang, Jie.Lu}@uts.edu.au |
| Pseudocode | Yes | Algorithm 1: LDD Drifted Instance Selection (LDD-DIS) |
| Open Source Code | Yes | The source code of LDD-DSDA is available at https://sites.google.com/view/anjin-concept-drift/home. |
| Open Datasets | Yes | Experiment 4 Electricity Price Prediction Data Set (Elec) Elec contains 45,312 instances, collected every thirty minutes from the Australian New South Wales Electricity Market between 7 May 1996 and 5 Dec 1998. Experiment 5 Nebraska Weather Prediction Data Set (Weather) This data set was compiled by the U.S. National Oceanic and Atmospheric Administration. Experiment 6 Spam Filtering Data Set (Spam Filtering) This data set is a collection of 9,324 email messages derived from the Spam Assassin collection. |
| Dataset Splits | No | The paper mentions '1,000 training cases' and '500 testing cases' for synthetic data, but does not explicitly detail a separate validation split or strategy for real-world datasets. |
| Hardware Specification | Yes | All experiments are conducted on a 2 3.1GHz 8 core CPU 128GB RAM cluster node with unique access. |
| Software Dependencies | No | All these algorithms were implemented based on the MOA framework [Bifet, et al., 2010], which is a commonly used software for evolving data stream analysis. However, no specific version numbers for MOA or other software dependencies are provided. |
| Experiment Setup | Yes | The configurations of LDD are set as the default values described in Algorithm 1. To fairly compare these drift adaptation algorithms, the default parameters suggested by the authors are used, and the base classification model is set as Naïve Bayes classifier, except for SAMk NN, which is only designed for IBk classifier. The parameters for IBk classifier of SAMk NN are 𝑘= 10 and weighting method = uniformly weighted. |