nTreeClus: a Tree-based Sequence Encoder for Clustering Categorical Series

TL;DR

nTreeClus introduces a novel tree-based sequence encoder for clustering categorical and time series data, demonstrating improved accuracy and robustness over existing methods through extensive empirical evaluation.

Contribution

It presents a new model-based clustering approach using tree learners and autoregressive models, with a unique numerical representation for categorical sequences.

Findings

Outperforms baseline methods in simulated scenarios.

Achieves up to 10.7% improvement in internal validation metrics.

Shows competitive or superior performance on real datasets.

Abstract

The overwhelming presence of categorical/sequential data in diverse domains emphasizes the importance of sequence mining. The challenging nature of sequences proves the need for continuing research to find a more accurate and faster approach providing a better understanding of their (dis)similarities. This paper proposes a new Model-based approach for clustering sequence data, namely nTreeClus. The proposed method deploys Tree-based Learners, k-mers, and autoregressive models for categorical time series, culminating with a novel numerical representation of the categorical sequences. Adopting this new representation, we cluster sequences, considering the inherent patterns in categorical time series. Accordingly, the model showed robustness to its parameter. Under different simulated scenarios, nTreeClus improved the baseline methods for various internal and external cluster validation metrics for up to 10.7% and 2.7%, respectively. The empirical evaluation using synthetic and real datasets, protein sequences, and categorical time series showed that nTreeClus is competitive or superior to most state-of-the-art algorithms.