A Broad Ensemble Learning System for Drifting Stream Classification
In a data stream environment, classification models must effectively and efficiently handle concept drift. Ensemble methods are widely used for this purpose; however, the ones available in the literature either use a large data chunk to update the model or learn the data one by one. In the former, the model may miss the changes in the data distribution, while in the latter, the model may suffer from inefficiency and instability. To address these issues, we introduce a novel ensemble approach based on the Broad Learning System (BLS), where mini chunks are used at each update. BLS is an effective lightweight neural architecture recently developed for incremental learning. Although it is fast, it requires huge data chunks for effective updates and is unable to handle dynamic changes observed in data streams. Our proposed approach, named Broad Ensemble Learning System (BELS), uses a novel updating method that significantly improves best-in-class model accuracy. It employs an ensemble of output layers to address the limitations of BLS and handle drifts. Our model tracks the changes in the accuracy of the ensemble components and reacts to these changes. We present the mathematical derivation of BELS, perform comprehensive experiments with 35 datasets that demonstrate the adaptability of our model to various drift types, and provide its hyperparameter, ablation, and imbalanced dataset performance analysis. The experimental results show that the proposed approach outperforms 10 state-of-the-art baselines, and supplies an overall improvement of 18.59% in terms of average prequential accuracy.
All the real and synthetic datasets are available in two formats : CSV and ARFF.
Google Drive Link: https://drive.google.com/drive/folders/16Nn9xmaMjbrR1btzN7__sSA5LXEgZyQc?usp=sharing
Python: 3.10.9
Numpy: 1.23.5
Pandas: 1.5.3
To execute the code, ensure that all the code files and the dataset (in .CSV format) are placed within the same folder. In the BELS_test.py file, make sure to include your dataset name using the "dataset_name" variable in the format: dataset_name = "YOUR_DATASET_NAME". After making this change, run the BELS_test.py file.
@ARTICLE{10225305,
author={Bakhshi, Sepehr and Ghahramanian, Pouya and Bonab, Hamed and Can, Fazli},
journal={IEEE Access},
title={A Broad Ensemble Learning System for Drifting Stream Classification},
year={2023},
volume={11},
number={},
pages={89315-89330},
doi={10.1109/ACCESS.2023.3306957}}