Text Categorization Based on Regularized Linear Classification Methods

By Tong Zhang, Frank J. Oles. Information retrieval (Boston), 2001

444 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Information Systems

View Paper: https://www.semanticscholar.org/paper/d49bfad00101a03ac3e964bea3717c75a5bb3210

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

In Information Systems

Citations in the corpus

Solving large scale linear prediction problems using stochastic gradient descent algorithms (2004, methodology): “Since regularized linear classi ers achieve state of the art performance in text-categorization (Li & Yang, 2003; Zhang & Oles, 2001), we use text data in our experiments.”
Proposal Scaling Distributed Machine Learning with System and Algorithm Co-design (2016, unknown): “In particular, we apply upper bounds of the diagonal entries of the Hessian for the coordinatespecific learning [154].”
PCCA: A new approach for distance learning from sparse pairwise constraints (2012, background): “where `β(x) = 1 β log(1 + e ) is the generalized logistic loss function [32].”
Uncovering shared structures in multiclass classification (2007, methodology): “Since neither the hinge []+ nor 10 the max operators are differentiable we employ an adaptation of the log-loss for the multiclass setting, with a parameter . controlling its sharpness (Zhang & Oles, 2001; Dekel et al., 2003), .”
Recursive regularization for large-scale classification with hierarchical and graphical dependencies (2013, background): “Moreover, most of the classes are linearly separable and finding a decision boundary is an ‘easy’ problem - this makes co-ordinate descent a natural choice for optimization in the context of text classification [13], [38].”
Large-Scale Bayesian Logistic Regression for Text Categorization (2007, methodology): “We based our implementation on the CLG algorithm (Zhang and Oles 2001), a cyclic coordinate descent (Luenberger 1984, sec.”
Feature selection for text categorization on imbalanced data (2004, methodology): “The micro averaged F1(BEP) score of LR is slightly lower than [17] due to a difference of treating document titles.”
Fast and simple gradient-based optimization for semi-supervised support vector machines (2014, methodology): “Here, the differentiable replacement for the hinge loss is the modified logistic loss [32]”
Stochastic Gradient Made Stable: A Manifold Propagation Approach for Large-Scale Optimization (2015, background): “However, as discovered in [22], [23], hinge loss can be smoothed by the loss of “modified logistic regression””
The MID4 dataset, 2002–2010: Procedures, coding rules and description (2015, methodology): “SVMs are commonly used algorithms for the classification of text documents (Abe, 2010; Dumais et al., 1998; Zhang and Oles, 2001).”
Person Re-Identification Using Kernel-Based Metric Learning Methods (2014, methodology): “To this effect, without loss of generality, they set T = 1 and then approximated the hinge loss with the generalized logistic loss function [27] β(x) = 1 β log(1 + e ) to form the objective function:”
A Comparison of Optimization Methods and Software for Large-scale L1-regularized Linear Classification (2010, methodology): “This method is extended from Zhang and Oles (2001) for L2-regularized logistic regression.”
Feature-Rich Part-of-Speech Tagging with a Cyclic Dependency Network (2003, background): “For example, Zhang and Oles (2001) attribute the perceived limited success of logistic regression for text categorization to a lack of use of regularization.”
A Quasi-Newton Approach to Nonsmooth Convex Optimization Problems in Machine Learning (2008, background): “Some machine learning applications have also been pursued along these lines (Chapelle, 2007; Zhang and Oles, 2001).”