Computational prediction of protein localization is one common way to characterize the functions of newly sequenced proteins. Sequence features such as amino acid (AA) composition have been widely used for subcellular localization prediction due to their simplicity while suffering from low coverage and low prediction accuracy. We present a physichemical encoding method that maps protein sequenc...

Most of the existing models for large scale arterial networks are not realistic and appropriate to deal with crowded conditions. As an alternative, we propose observation-based models that circumvent the fragility problems of traditional models. Monitoring replaces prediction, and the system is repeatedly modified based on observations. To succeed this goal a city is modeled in an aggregated ma...

We present a self consistent method to perfom a joint analysis of Sunyaev-Zel’dovich and weak gravitational lensing observation of galaxy clusters. The spatial distribution of the cluster main constituents is described by a perturbative approach. Assuming the hydrostatic equilibrium and the equation of state, we are able to deduce, from observations, maps of projected gas density and gas temper...

Existing computational models of visual attention generally employ simple image features such as color, intensity or orientation to generate a saliency map which highlights the image parts that attract human attention. Interestingly, most of these models do not process any depth information and operate only on standard two-dimensional RGB images. On the other hand, depth processing through ster...

Recurrent Self-Organizing Map (RSOM) is studied in temporal sequence processing. RSOM includes a recurrent difference vector in each unit of the map, which allows storing temporal context from consecutive input vectors fed to the map. RSOM is a modification of the Temporal Kohonen Map (TKM). It is shown that RSOM learns a correct mapping from temporal sequences of a simple synthetic data, while...

Structured prediction is the problem of learning a function that maps structured inputs to structured outputs. Prototypical examples of structured prediction include part-ofspeech tagging and semantic segmentation of images. Inspired by the recent successes of search-based structured prediction, we introduce a new framework for structured prediction called HC-Search. Given a structured input, t...

This papC'r describes research conducted towards Columbia's unified Protocol Implemelltation and Design (CUPID) environment. CUPID research aims at the integration and automation of protocol design and implementation tools. C{ -PID uses an algebraic representation of protocols based. in part. upon a variant of \lilner's calculus of communicating systems ICCS). Communication behaviors are repres...

Let φ be a function that maps any non-empty subset A of R to a non-empty subset φ(A) of R. A φ-cover of a set T = {T1, T2, . . . , Tm} of pairwise non-crossing trees in the plane is a set of pairwise disjoint connected regions such that 1. each tree Ti is contained in some region of the cover, 2. each region of the cover is either (a) φ(Ti) for some i, or (b) φ(A ∪B), where A and B are construc...

The 24-hour 500-mb. barotropic forrcasts prepared by the Joint Numerical Weather Prediction Unit (JNWPU) have been investigated in 30 cases of rapid sea level cyclogenesis. Composite error maps are presented for the region of cyclogenesis. The 500-mb. errors are found t o bc significantly larger when the solenoidal field at that level is strong than when it is weak.

In this paper, we develop segmental semi-Markov models (SSMM) to exploit alignment profiles for protein secondary structure prediction. A novel parameterized model is proposed as the likelihood function for the SSMM to capture the segmental conformation from the profiles. By incorporating the information of long range interactions in β-sheets, this model is capable to carry out inference on con...