RadarSeq: A Temporal Vision Framework for User Churn Prediction via Radar Chart Sequences

TL;DR

RadarSeq introduces a novel temporally-aware computer vision framework that models user behavior as radar chart sequences, effectively predicting churn in gig platforms by capturing spatial-temporal patterns and outperforming existing methods.

Contribution

The paper presents a new radar chart sequence modeling approach combining CNN and LSTM for user churn prediction, enhancing interpretability and performance over classical and ViT-based baselines.

Findings

Achieved 17.7% F1 score improvement

Improved precision by 29.4%

Enhanced AUC by 16.1%

Abstract

Predicting user churn in non-subscription gig platforms, where disengagement is implicit, poses unique challenges due to the absence of explicit labels and the dynamic nature of user behavior. Existing methods often rely on aggregated snapshots or static visual representations, which obscure temporal cues critical for early detection. In this work, we propose a temporally-aware computer vision framework that models user behavioral patterns as a sequence of radar chart images, each encoding day-level behavioral features. By integrating a pretrained CNN encoder with a bidirectional LSTM, our architecture captures both spatial and temporal patterns underlying churn behavior. Extensive experiments on a large real-world dataset demonstrate that our method outperforms classical models and ViT-based radar chart baselines, yielding gains of 17.7 in F1 score, 29.4 in precision, and 16.1 in AUC, along with improved interpretability. The framework's modular design, explainability tools, and efficient deployment characteristics make it suitable for large-scale churn modeling in dynamic gig-economy platforms.