Zero-shot learning has gained popularity due to its potential to scale recognition models without requiring additional training data. This is usually achieved by associating categories with their semantic information like attributes. However, we believe that the potential offered by this paradigm is not yet fully exploited. In this work, we propose to utilize the structure of the space spanned ...

The embedding dimension and the number of nearest neighbors are very important parameters in the prediction of a chaotic time series. In order to reduce the uncertainties in the determination of the forgoing two parameters, a new adaptive local linear prediction method is proposed in this study. In the new method, the embedding dimension and the number of nearest neighbors are combined as a par...

This paper considers the problem of blind active spreadspectrum (SS) steganalysis defined as the extraction of hidden data with no prior information. We first develop a multisignature iterative generalized least-squares (M-IGLS) core procedure to seek unknown messages hidden in image hosts via multi-signature direct-sequence spread-spectrum embedding. Neither the original host nor the embedding...

Hybrid QM:QM (quantum mechanics:quantum mechanics) and QM:MM (quantum mechanics:molecular mechanics) methods are widely used to calculate the electronic structure of large systems where a full quantum mechanical treatment at a desired high level of theory is computationally prohibitive. The ONIOM (our own N-layer integrated molecular orbital molecular mechanics) approximation is one of the more...

In 1998 Crandall introduced a method based on coding theory to secretly embed a message in a digital support such as an image. Later Fridrich et al. improved this method to minimize the distortion introduced by the embedding; a process called wet paper. However, as previously emphasized in the literature, this method can fail during the embedding step. Here we find sufficient and necessary cond...

We present a new class of metrics on R, which we call the scaled embedding-generated (SEG) metric. These metrics are defined with the help of a smooth one-to-one embedding of R into R and an additional scalar-valued function which is used to re-scale the distances. We describe a possible construction of an SEG metric which is based on the GaußNewton algorithm. More precisely, we show how to gen...

Abstract. We develop a theory of Lp spaces based on outer measures generated through coverings by distinguished sets. The theory includes as a special case the classical Lp theory on Euclidean spaces as well as some previously considered generalizations. The theory is a framework to describe aspects of singular integral theory, such as Carleson embedding theorems, paraproduct estimates, and T (...

Manifold learning is a powerful tool for reducing the dimensionality of a dataset by finding a low-dimensional embedding that retains important geometric and topological features. In many applications it is desirable to add new samples to a previously learnt embedding, this process of adding new samples is known as the out-ofsample extension problem. Since many manifold learning algorithms do n...

We extend existing question answering (QA) system to deal with words that are unseen or do not have enough examples during training. Instead of learning word embedding from scratch for a specific QA dataset, we decompose the embedding into two components: The first captures the general semantic meaning of a word and can be trained using readily available large corpuses. The second component rew...

Current applications of frozen-density embedding (FDE)—or more generally subsystem density-functional theory (DFT) schemes—are limited to subsystems that are not connected by covalent bonds. This restriction is due to the insufficiencies of the available approximate kinetic-energy functionals, which are used to calculate the contribution of the nonadditive kinetic energy to the effective embedd...