Zero-shot learning has received increasing interest as a means to alleviate the often prohibitive expense of annotating training data for large scale recognition problems. These methods have achieved great success via learning intermediate semantic representations in the form of attributes and more recently, semantic word vectors. However, they have thus far been constrained to the single-label...

There is recently a surge in approaches that learn low-dimensional embeddings of nodes networks. However, for large-scale real-world networks, it’s inefficient existing to store amounts parameters memory and update them edge by edge. With the knowledge having similar neighborhoods will be close each other embedding space, we propose COSINE (COmpresSIve Network Embedding) algorithm, which reduce...

label switching technology is a flexible and high performance method for forwarding layer three packets that are mapped to layer two flows. each label switch router, lsr, needs a specific label for every flow, where the number of labels depends on the mapping and granularity of flows. data and control driven methods are the most significant and popular methods in label mapping policy. the above...

Conventional multi-label classification algorithms treat the target labels of the classification task as mere symbols that are void of an inherent semantics. However, in many cases textual descriptions of these labels are available or can be easily constructed from public document sources such as Wikipedia. In this paper, we investigate an approach for embedding documents and labels into a join...

The key to semi-supervised learning (SSL) is explore adequate information leverage the unlabeled data. Current dominant approaches aim generate pseudo-labels on weakly augmented instances and train models their corresponding strongly variants with high-confidence results. However, such methods are limited in excluding samples low-confidence under-utilization of label information. In this paper,...

Recently, zero-shot learning (ZSL) has received increasing interest. The key idea underpinning existing ZSL approaches is to exploit knowledge transfer via an intermediate-level semantic representation which is assumed to be shared between the auxiliary/source dataset and the target/test dataset and re-used as a bridge between the source and target domains for knowledge transfer. The semantic r...

In a traditional setting, classifiers are trained to approximate a target function f : X → Y where at least a sample for each y ∈ Y is presented to the training algorithm. In a zero-shot setting we have a subset of the labels Ŷ ⊂ Y for which we do not observe any corresponding training instance. Still, the function f that we train must be able to correctly assign labels also on Ŷ . In practice,...

Abstract Graph representation learning methods opened new avenues for addressing complex, real-world problems represented by graphs. However, many graphs used in these applications comprise millions of nodes and billions edges are beyond the capabilities current software implementations. We present GRAPE (Graph Representation Learning, Prediction Evaluation), a resource graph processing embeddi...

‎for a given measure space $(x,{mathscr b},mu)$ we construct all measure spaces $(y,{mathscr c},lambda)$ in which $(x,{mathscr b},mu)$ is embeddable‎. ‎the construction is modeled on the ultrafilter construction of the stone--v{c}ech compactification of a completely regular topological space‎. ‎under certain conditions the construction simplifies‎. ‎examples are given when this simplification o...

Multi-label learning aims to automatically assign to an instance (e.g., an image or a document) the most relevant subset of labels from a large set of possible labels. The main challenge is to maintain accurate predictions while scaling efficiently on data sets with extremely large label sets and many training data points. We propose a simple but effective neural net approach, the Semantic Embe...