HOT-VAE: Learning High-Order Label Correlation for Multi-Label Classification via Attention-Based Variational Autoencoders

Abstract

Understanding how environmental characteristics affect biodiversity patterns, from individual species to communities of species, is critical for mitigating effects of global change. A central goal for conservation planning and monitoring is the ability to accurately predict the occurrence of species communities and how these communities change over space and time. This in turn leads to a challenging and long-standing problem in the field of computer science -how to perform accurate multi-label classification with hundreds of labels? The key challenge of this problem is its exponential-sized output space with regards to the number of labels to be predicted. Therefore, it is essential to facilitate the learning process by exploiting correlations (or dependency) among labels. Previous methods mostly focus on modelling the correlation on label pairs; however, complex relations between real-world objects often go beyond second order. In this paper, we propose a novel framework for multi-label classification, Highorder Tie-in Variational Autoencoder (HOT-VAE), which performs adaptive high-order label correlation learning. We experimentally verify that our model outperforms the existing state-of-the-art approaches on a bird distribution dataset on both conventional F1 scores and a variety of ecological metrics. To show our method is general, we also perform empirical analysis on seven other public real-world datasets in several application domains, and Hot-VAE exhibits superior performance to previous methods.

Deep learning with citizen science data enables estimation of species diversity and composition at continental extents

“…These advances greatly expand the scope of inference, making it feasible to predict, document, and study patterns of species diversity and community composition across a broader range of spatial and temporal scales. The application of deep learning methods that estimate relative abundance for multiple species (Kong et al, 2020), accommodate other data types (e.g., time series data), or that relax the assumption of symmetric, pairwise associations (Zhao et al, 2021) between species may also lead to more accurate predictions of species diversity and composition that can better inform the prioritization of limited conservation resources (Johnston et al, 2015). ECOLOGY Addressing challenges associated with the interpretability of deep learning models and, more specifically, prediction uncertainty can further inform conservation decision-making and ensure that resources are more precisely directed to critical areas associated with high prediction certainty (Jansen et al, 2022;Wadoux et al, 2023).…”

Multi-relation Message Passing for Multi-label Text Classification

“…The label graph can be combined with graph neural networks to enable the interactive learning of features and label embeddings. [5] and [18] propose starting with a simple fully-connected graph or a co-occurrence label graph. A Message Passing Neural Network (MPNN) then passes messages among label embeddings and features to enable learning of high order label correlation structure that is conditioned on the features.…”