Discriminative Transfer Learning with Tree-based Priors

By Nitish Srivastava, Ruslan Salakhutdinov. 2013

223 citations recorded by the source corpus.

Subfields: Representation & Self-Supervised Learning

How researchers see this paper

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

In Representation & Self-Supervised Learning

Citations in the corpus

Deeply-Supervised Nets (2014, methodology): “…dataset, run one forward pass, and show the feature maps learned from the ﬁrst (bottom) convolutional layer in Figure (3) [32] 42.51 Maxout Networks [9] 38.57 Tree based Priors [27] 36.85 Network in Network [20] CIFAR-100 dataset is similar to CIFAR-10 dataset, except that it has 100 classes.”
Striving for Simplicity: The All Convolutional Net (2014, methodology): “Among these the most notable directions are work on using more complex activation functions (Goodfellow et al., 2013; Lin et al., 2014; Srivastava et al., 2013) techniques for improving class inference (Stollenga et al., 2014; Srivastava & Salakhutdinov, 2013) as well as procedures for improved…”
Recent advances in convolutional neural networks (2015, background): “The work of [94] is one of the earliest attempts to introduce category hierarchy in CNN, in which a discriminative”
Deep Learning: Methods and Applications (2014, background): “Srivastava and Salakhutdinov [349] present a multi-task learning approach based on hierarchical Bayesian priors in a DNN system applied to various image classification data sets.”
Discriminative k-shot learning using probabilistic models (2017, methodology): “[6] propose an elegant approach to k-shot learning that is an instance of the framework described here: a Gaussian model is fit to the weights with MAP inference.”
Generalizing Pooling Functions in Convolutional Neural Networks: Mixed, Gated, and Tree (2015, neutral): “Tree Priors [36] - - 36.”
Tree-CNN: A hierarchical Deep Convolutional Neural Network for incremental learning (2018, background): ““Discriminative Transfer learning” [28] is one of the earliest works where classes are categorized hierarchically to improve network performance.”
A Review of Domain Adaptation without Target Labels (2019, methodology): “using decision-tree priors [73], [201] and through hierarchical priors [195], [197].”
Maximum a posteriori adaptation of network parameters in deep models (2015, background): “13 over θs with fixed DNN weights (see [38]).”
Beyond One-hot Encoding: lower dimensional target embedding (2018, methodology): “In the ﬁeld of large-scale recognition, hierarchical approaches such as using tree-based priors [38], label relational graphs [11], CNN hierarchies [46], and HD-CNNs [47] have been proposed.”
Fine-Grained Image Classification by Exploring Bipartite-Graph Labels (2015, methodology): “Compared to [47], our method is able to handle more general structure (e.”
A Deep and Autoregressive Approach for Topic Modeling of Multimodal Data (2014, methodology): “The RS model was later used for multimodal data modeling [12], where pairs of images and text annotations were modeled jointly within a deep Boltzmann machine (DBM) [13].”
DisturbLabel: Regularizing CNN on the Loss Layer (2016, unknown): “81 Srivastava [32] 36.”
Deep Convolutional Neural Networks for Image Classification: A Comprehensive Review (2017, background): “…Wan, Zeiler, Zhang, LeCun, & Fergus, 2013; Goodfellow, Warde-Farley, Mirza, Courville, & Bengio, 2013) and CIFAR-100 benchmarks (Krizhevsky, 2009; Srivastava & Salakhutdinov, 2013), and very close to the state of the art on the MNIST 2 Abstract in this contents relates to features that are…”
Reconciling meta-learning and continual learning with online mixtures of tasks (2018, methodology): “This allows our algorithm to make use of a scalable stochastic gradient descent optimizer instead of alternating a maximization step with an inference pass over the full dataset [ c.f., 50, 3].”