Learn More
We introduce novel profile-based string kernels for use with support vector machines (SVMs) for the problems of protein classification and remote homology detection. These kernels use probabilistic profiles, such as those produced by the PSI-BLAST algorithm, to define position-dependent mutation neighborhoods along protein sequences for inexact matching of(More)
  • Pe T E R K N Ippe Rt Z, E Rt A N Sm A N N, D Ie, T R Ic, H A Lt, H Au Se N +18 others
  • 2009
A B S T R A C T The SAMUM field campaign in southern Morocco in May/June 2006 provides valuable data to study the emission, and the horizontal and vertical transports of mineral dust in the Northern Sahara. Radiosonde and lidar observations show differential advection of air masses with different characteristics during stable nighttime conditions and up to(More)
  • E Ik, E B Ie, Rw Irt H, N Fr E D W E N D Isc H, A N D R ´ E Eh Rlich, B Ir +20 others
  • 2009
A B S T R A C T In May–June 2006, airborne and ground-based solar (0.3–2.2 μm) and thermal infrared (4–42 μm) radiation measurements have been performed in Morocco within the Saharan Mineral Dust Experiment (SAMUM). Upwelling and downwelling solar irradiances have been measured using the Spectral Modular Airborne Radiation Measurement System(More)
In content-based audio retrieval, the goal is to find sound recordings (audio documents) based on their acoustic features. This content-based approach differs from retrieval approaches that index media files using metadata such as file names and user tags. In this paper, we propose a machine learning approach for retrieving sounds that is novel in that it(More)
Predicting a protein's structural class from its amino acid sequence is a fundamental problem in computational biology. Recent machine learning work in this domain has focused on developing new input space representations for protein sequences, that is, string kernels, some of which give state-of-the-art performance for the binary prediction task of(More)
We develop a novel multi-class classification method based on <i>output codes</i> for the problem of classifying a sequence of amino acids into one of many known protein structural classes, called <i>folds.</i> Our method learns relative weights between one-vs-all classifiers and encodes information about the protein structural hierarchy for multi-class(More)
BACKGROUND Predicting a protein's structural class from its amino acid sequence is a fundamental problem in computational biology. Much recent work has focused on developing new representations for protein sequences, called string kernels, for use with support vector machine (SVM) classifiers. However, while some of these approaches exhibit state-of-the-art(More)
  • R A L Ph K A H N, N Fr E D W E N D Isc H, E Ik, E B Ie, Rw Irt H, T Il M A N D In T E R +1 other
  • 2009
A B S T R A C T Coincident observations made over the Moroccan desert during the Sahara mineral dust experiment (SAMUM) 2006 field campaign are used both to validate aerosol amount and type retrieved from multi-angle imaging spectroradiometer (MISR) observations, and to place the suborbital aerosol measurements into the satellite's larger regional context.(More)
Object recognition and localization are important tasks in computer vision. The focus of this work is the incorporation of contextual information in order to improve object recognition and localization. For instance, it is natural to expect not to see an elephant to appear in the middle of an ocean. We consider a simple approach to encapsulate such common(More)