Subgradient Methods for Structured Prediction

By Nathan D. Ratliff. 2007

How researchers see this paper

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

Citations in the corpus

Fastfood: Approximate Kernel Expansions in Loglinear Time (2014, background): “Recent advances in both online (Ratliﬀ et al., 2007) and batch (Teo et al., 2010; Boyd et al., 2010)”
Learning-Based Single-Document Summarization with Compression and Anaphoricity Constraints (2016, methodology): “We train the model via stochastic subgradient descent on the primal form of the structured SVM objective (Ratliff et al., 2007; Kummerfeld et al., 2015).”
Predicting Program Properties from "Big Code" (2015, methodology): “In particular, we use a variant of the Structured Support Vector Machine (SSVM) 5 [27] and we train it efﬁciently with the scalable subgradient descent algorithm proposed in [23].”
Parsing with Compositional Vector Grammars (2013, methodology): “This max-margin, structure-prediction objective (Taskar et al., 2004; Ratliff et al., 2007; Socher et al., 2011b) trains the CVG so that the highest scoring tree will be the correct tree: g θ ( x i ) = y i and its score will be larger up to a margin to other possible trees ˆ y ∈ Y ( x i ) :”
A Reduction of Imitation Learning and Structured Prediction to No-Regret Online Learning (2010, methodology): “The k th output binary variable ˆ y k = I ( w Tk x + b k > 0) , where w k , b k optimizes the SVM objective with regularizer λ = 10 − 4 using stochastic gradient descent (Ratliff et al., 2007; Bottou, 2009).”
Discriminative Unsupervised Alignment of Natural Language Instructions with Corresponding Video Segments (2015, methodology): “LSSVM minimizes the following objective function: which is non-convex and non-differentiable, and optimized utilizing the subgradient method (Ratliff et al., 2007).”
Max-Margin Tensor Neural Network for Chinese Word Segmentation (2014, methodology): “We use a generalization of gradient descent called subgradient method (Ratliff et al., 2007) which computes a gradient-like direction.”
CHOMP: Covariant Hamiltonian optimization for motion planning (2013, background): “Under these conditions, the overall CHOMP objective function is strongly convex – that is, it can be lower-bounded over the entire region by a quadratic with curvature A [64].”
Cutting-plane training of structural SVMs (2009, methodology): “Such algorithms either take stochastic subgradient steps (Collins 2002; Ratliff et al. 2007; Shalev-Shwartz et al. 2007), or build a cutting-plane model which is easy to solve directly (Tsochantaridis et al. 2004).”
Learning preferences for manipulation tasks from online coactive feedback (2015, background): “In Ratliff et al. [45] the robot observes optimal user feedback but performs approximate inference.”