Biomedical subjects

M Horoi

Publications and source records attributed to M Horoi.

13 recordsLinked to original sources

PubMed2001

Random versus realistic interactions for low-lying nuclear spectra.

We compare the shell-model results for realistic interactions with those obtained for various ensembles of random matrix elements. We show that, although the quantum numbers of the ground states in the even-even nuclei have a high probability ( approximately 60%) to be J(pi)T = 0(+)0, the overlap of those states with the realistic wave functions is very small in average. The transition probabilities B(E2) predicted with random interactions are also too small. The presence of the regular pairing is shown to be a significant element of realistic physics not reproduced by random interactions.

Journal Article↗

PubMed2001

Spectral statistics of the two-body random ensemble revisited.

Using longer spectra we reanalyze spectral properties of the two-body random ensemble studied 30 years ago. At the center of the spectra the old results are largely confirmed, and we show that the nonergodicity is essentially due to the variance of the lowest moments of the spectra. The longer spectra allow us to test and reach the limits of validity of French's correction for the number variance. At the edge of the spectra we discuss the problems of unfolding in more detail. With a Gaussian unfolding of each spectrum the nearest-neighbor spacing distribution between ground state and first exited state is shown to be stable. Using such an unfolding the distribution tends toward a semi-Poisson distribution for longer spectra. For comparison with the nuclear table ensemble we could use such unfolding obtaining similar results as in the early papers, but an ensemble with realistic splitting gives reasonable results if we just normalize the spacings in accordance with the procedure used for the data.

Journal Article↗

PubMed2000

Single-parent evolution algorithm and the optimization of Si clusters.

We describe a novel method for the structural optimization of molecular systems. Similar to genetic algorithms (GA), our approach involves an evolving population in which new members are formed by cutting and pasting operations on existing members. Unlike previous GA's, however, the population in each generation has a single parent only. This scheme has been used to optimize Si clusters with 13-23 atoms. We have found a number of new isomers that are lower in energy than any previously reported and have properties in much better agreement with experimental data.

Journal Article↗

PubMed1995