arXiv subjects
Publications and source records attributed to Xiao Pu.
Consider the problem of sparse clustering, where it is assumed that only a subset of the features are useful for clustering purposes. In the framework of the COSA method of Friedman and Meulman, subsequently improved in the form of the Sparse K-means method of Witten and Tibshirani, a natural and simpler hill-climbing approach is introduced. The new method is shown to be competitive with these two methods and others.
We adapt the compression algorithm of Weinstein, Auerbach, and Chandra from eigenvectors of spin lattice Hamiltonians to eigenvectors of light-front field-theoretic Hamiltonians. The latter are approximated by the standard discrete light-cone quantization technique, which provides a matrix representation of the Hamiltonian eigenvalue problem. The eigenvectors are represented as singular value decompositions of two-dimensional arrays, indexed by transverse and longitudinal momenta, and compressed by truncation of the decomposition. The Hamiltonian is represented by a rank-four tensor that is decomposed as a sum of contributions factorized into direct products of separate matrices for transverse and longitudinal interactions. The algorithm is applied to a model theory, to illustrate its use.