FastXML: a fast, accurate and stable tree-classifier for extreme multi-label learning

By Yashoteja Prabhu, Manik Varma. 2014

423 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

Citations in the corpus

Deep Metric Learning via Lifted Structured Feature Embedding (2015, background): “Deep metric learning: Bromley et al. [3] paved the way on deep metric learning and trained Siamese networks for signature verification.”
Neural Message Passing for Multi-Label Classification (2019, background): “Reuters Bibtex Bookmarks Delicious RCV1 TFBS SIDER NUSWIDE FastXML [37] - - 0.”
The Emerging Trends of Multi-Label Learning (2020, background): “FastXML, PfastreXML and SwiftXML have studied the ranking-based measures such as nDCG and its variants.”
Deep Extreme Multi-label Learning (2017, methodology): “abel random forest method (MLRF) [1] seeks to learn an ensemble of randomized trees instead of relying on a fundamental classifier. MLRF also suffers the extremely high training cost as LPSR. FastXML [21], which is considered the state-of-the-art tree based method for XML, is proposed to learn the h”
Towards Scalable and Reliable Capsule Networks for Challenging NLP Applications (2019, methodology): “Baselines We compare our approach to the following baselines: non-deep learning approaches using TF-IDF features of documents as inputs: FastXML (Prabhu and Varma, 2014), and PD-Sparse (Yen et al., 2016), deep learning approaches using raw text of documents as inputs: FastText (Joulin et al.,…”
Learning Tree-based Deep Model for Recommender Systems (2018, background): “Unlikemulticlass andmulti-label classification works [26, 32] where tree is used to partition samples or labels, our recommendation tree indexes items for retrieval.”
Weakly-Supervised Hierarchical Text Classification (2018, background): “FastXML (Prabhu and Varma 2014) is designed for extremely large label space.”
AttentionXML: Label Tree-based Attention-Aware Deep Model for High-Performance Extreme Multi-Label Text Classification (2018, neutral): “They can be categorized into the following four types: 1) 1-vs-All [3,4,30,31], 2) Embedding-based [7,27], 3) Instance [10,25] or label tree-based [11, 13, 24, 28]) and 4) Deep learning-based methods [17] (see Appendix for more descriptions on these methods).”
Sparse Local Embeddings for Extreme Multi-label Classification (2015, methodology): “For instance, FastXML [15] can achieve a prediction accuracy of 49% on WikiLSHTC using a 50 tree ensemble.”
Deep Metric Learning via Facility Location (2016, background): “Furthermore, metric learning can be used for challenging extreme classification settings [23, 5, 40], where the number of classes is very large and the number of examples per class becomes scarce.”
Probabilistic Label Tree for Streaming Multi-Label Learning (2021, background): “There are typically two types of trees including instance trees [18, 21] and label trees [3, 10, 32], depending whether instance or label is partitioned in tree nodes.”
Deep Learning for Extreme Multi-label Text Classification (2017, methodology): “These methods mainly differ in their choices of compression and decompression techniques[36] such as compressed sensing[19, 24], Bloom filters[10], Singular Value Decomposition[39], landmark labels[3], output codes[50], etc.”
Deep Streaming Label Learning (2020, background): “…(LE) is a popular strategy for MLL that projects the label vectors onto a lower subspace with the latent embedding of corresponding information and utilizes a decompression mapping to lift the embedded label vectors back to the original label space (Tsoumakas et al., 2011; Prabhu & Varma, 2014).”
Contextual Memory Trees (2018, methodology): “Hence we observe for RCV1-1K with 1K labels, OAA can actually be more computationally efficient then CMT.”
Maximizing Subset Accuracy with Recurrent Neural Networks in Multi-label Classification (2017, methodology): “The margin of FastXML to RNNm and EncDec is further increased.”