Deep Attention-Based Supernovae Classification of Multi-Band Light-Curves

TL;DR

This paper introduces a deep attention-based model for classifying supernovae from irregular multi-band light-curves, leveraging synthetic data for training and demonstrating improved accuracy and interpretability over existing methods.

Contribution

The paper presents a novel deep attention model that classifies supernovae without feature engineering, uses synthetic data for pre-training, and provides interpretability insights into the model's focus.

Findings

Outperformed RNN-based models in late and early classification scenarios.

Achieved higher balanced-$F_1$ score compared to traditional classifiers.

Synthetic data pre-training enhanced model performance and robustness.

Abstract

In astronomical surveys, such as the Zwicky Transient Facility, supernovae (SNe) are relatively uncommon objects compared to other classes of variable events. Along with this scarcity, the processing of multi-band light-curves is a challenging task due to the highly irregular cadence, long time gaps, missing-values, few observations, etc. These issues are particularly detrimental to the analysis of transient events: SN-like light-curves. We offer three main contributions: 1) Based on temporal modulation and attention mechanisms, we propose a Deep attention model (TimeModAttn) to classify multi-band light-curves of different SN types, avoiding photometric or hand-crafted feature computations, missing-value assumptions, and explicit imputation/interpolation methods. 2) We propose a model for the synthetic generation of SN multi-band light-curves based on the Supernova Parametric Model, allowing us to increase the number of samples and the diversity of cadence. Thus, the TimeModAttn model is first pre-trained using synthetic light-curves. Then, a fine-tuning process is performed. The TimeModAttn model outperformed other Deep Learning models, based on Recurrent Neural Networks, in two scenarios: late-classification and early-classification. Also, the TimeModAttn model outperformed a Balanced Random Forest (BRF) classifier (trained with real data), increasing the balanced-$F_1$score from $\approx.525$ to $\approx.596$. When training the BRF with synthetic data, this model achieved similar performance to the TimeModAttn model proposed while still maintaining extra advantages. 3) We conducted interpretability experiments. High attention scores were obtained for observations earlier than and close to the SN brightness peaks. This also correlated with an early highly variability of the learned temporal modulation.