Online Passive-Aggressive Algorithms

By Koby Crammer, Ofer Dekel, Joseph Keshet, Shai Shalev-Shwartz, Yoram Singer. 2003

2,132 citations recorded by the source corpus.

Subfields: Optimization & Training, Statistical Learning Theory

How researchers see this paper

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

In Optimization & Training

In Statistical Learning Theory

Citations in the corpus

Can Decentralized Algorithms Outperform Centralized Algorithms? A Case Study for Decentralized Parallel Stochastic Gradient Descent (2017, neutral): “SGD is closely related to online learning algorithms, for example, Crammer et al. [2006], Shalev-Shwartz [2011], Yang et al. [2014].”
Asynchronous Parallel Stochastic Gradient for Nonconvex Optimization (2015, methodology): “entially achieves the same convergence rate [Xiao, 2009]. The stochastic gradient is also closely related to online learning algorithms. Please refer to the online learning literatures, for example, [Crammer et al., 2006, Shalev-Shwartz, 2011, Yang et al., 2014]. Most studies for stochastic gradient”
Continual Few-Shot Learning for Text Classification (2021, background): “Online learning algorithms learn to update models from data streams sequentially, where the task is the same but can exhibit concept drift (new patterns) (Zinkevich, 2003; Crammer et al., 2006; Sahoo et al., 2018; Jerfel et al., 2019; Javed and White, 2019”
Towards Open World Recognition (2014, methodology): “While there is a significant body of work on incremental learning algorithms that handle new instances of known classes [3, 4, 45], open world requires two more general and difficult steps: continuously detecting novel classes and when novel inputs are found updating the system to include these new ”
Domain Adaptation with Structural Correspondence Learning (2006, methodology): “MIRA has been used successfully for both sequence analysis (McDonald et al., 2005a) and dependency parsing (McDonald et al., 2005b).”
Memory and Communication Efficient Distributed Stochastic Optimization with Minibatch Prox (2017, methodology): “This can be viewed as a minibatch generalization to the passive-aggressive algorithm (Crammer et al., 2006) and has been considered in various contexts (Kulis and Bartlett, 2010; Toulis and Airoldi, 2014).”
Learning Cumulatively to Become More Knowledgeable (2016, background): “Online learning and incremental learning [1, 6, 10, 15, 20, 42] mainly aim at handling new instances of known classes.”
11,001 New Features for Statistical Machine Translation (2009, methodology): “We incorporate all our new features into a linear model (Och and Ney, 2002) and train them using MIRA (Crammer et al., 2006), following previous work (Watanabe et al., 2007; Chiang et al., 2008).”
A Survey on Metric Learning for Feature Vectors and Structured Data (2013, methodology): “are able to optimize KM in an online manner using a simple and efficient algorithm, which belongs to the family of Passive-Aggressive algorithms (Crammer et al., 2006).”
Recent Advances in Open Set Recognition: A Survey (2018, background): “On the other hand, though new classes (UUCs) are assumed to appear incremental in class incremental learning (C-IL) [125], [126], [127], [128], [129], [130], these studies mainly focused on how to enable the system to incorporate later coming training samples from new classes instead of handling the”
Online Deep Metric Learning (2018, methodology): “Given a triplet 〈xt ,xp ,xq〉 at the t-th time step, inspired by PassiveAggressive (PA) [4] and OPML [16], we design a convex objective”
Communication-Efficient Distributed Online Prediction by Dynamic Model Synchronization (2014, background): “[6] also gives other variants of passive aggressive updates that have a reduced learning rate determined by an aggressiveness parameter C > 0.”
Label Embedding Trees for Large Multi-Class Tasks (2010, positive): “We also compare to three other baselines: One-vs-Rest large margin classifiers trained using the passive aggressive algorithm [9], the Filter Tree [7] and the Conditional Probability Tree (CPT) [6].”
The Statistics of Streaming Sparse Regression (2014, background): “[16] Koby Crammer, Ofer Dekel, Joseph Keshet, Shai Shalev-Shwartz, and Yoram Singer.”
Have You Lost the Thread? Discovering Ongoing Conversations in Scattered Dialog Blocks (2017, methodology): “Unfortunately, due to the stressing size of the DOC corpus, we have to use linear online learning models (Rosenblatt 1958; Crammer et al. 2006), which may result in a severe limitation in the expressiveness of feature spaces.”
Online Metric Learning for Multi-Label Classification (2020, methodology): “Loss Bound Following the analysis in (Crammer et al. 2006), we state the upper bounds for our online metric learning algorithm.”
A Discriminative Kernel-Based Approach to Rank Images from Text Queries (2008, methodology): “Moreover, PAMIR also enjoys an efficient learning algorithm which builds upon recent work on the online learning of kernel-based classifiers [12].”
Malicious URL Detection using Machine Learning: A Survey (2017, methodology): “Lastly, we discuss other learning algorithms, in challenges speciﬁc to Malicious URL Detection are addressed, including cost-sensitive learning, active learning, similarity learning, unsupervised learning and string pattern matching.”
NRC-Canada-2014: Detecting Aspects and Sentiment in Customer Reviews (2014, methodology): “The tagger was trained using the structured PassiveAggressive (PA) algorithm with a maximum stepsize of C = 1 (Crammer et al., 2006).”