本文選題：蛋白質(zhì)　切入點(diǎn)：模型質(zhì)量　出處：《計(jì)算機(jī)應(yīng)用研究》2017年05期 　論文類型：期刊論文

【摘要】：針對(duì)傳統(tǒng)蛋白質(zhì)模型質(zhì)量評(píng)估沒(méi)有考慮同源信息的問(wèn)題,提出了一種基于LS-SVM評(píng)估蛋白質(zhì)模型質(zhì)量的方法。綜合模擬退火(simulated annealing,SA)算法跳出局部最優(yōu)解和粒子群(particle swarm optimization,PSO)算法收斂速度快的特點(diǎn),提出了模擬退火粒子群(SAPSO)算法。利用SAPSO算法來(lái)優(yōu)化LS-SVM參數(shù)C和γ,最后得到最優(yōu)模型來(lái)評(píng)估蛋白質(zhì)模型質(zhì)量。實(shí)驗(yàn)結(jié)果表明,經(jīng)SAPSO優(yōu)化LS-SVM參數(shù)所得到的模型評(píng)估預(yù)測(cè)誤差較小,且預(yù)測(cè)值更穩(wěn)定。
[Abstract]:In view of the problem that the traditional protein model quality assessment does not consider the homology information, A method for evaluating the quality of protein model based on LS-SVM is proposed. The algorithm of synthetic simulated annealing and particle swarm optimization (PSOs) can jump out of the local optimal solution and the particle swarm optimization (PSOs) algorithm converges quickly. A simulated annealing particle swarm optimization (SAPSO) algorithm is proposed. SAPSO algorithm is used to optimize the parameters C and 緯 of LS-SVM, and the optimal model is obtained to evaluate the quality of protein model. The experimental results show that the prediction error of model evaluation obtained by SAPSO optimization of LS-SVM parameters is small. And the predicted value is more stable.
【作者單位】： 河南師范大學(xué)計(jì)算機(jī)與信息工程學(xué)院;
【基金】：國(guó)家自然科學(xué)基金資助項(xiàng)目(61173071) 河南省高校創(chuàng)新人才支持計(jì)劃項(xiàng)目(2012HASTIT011)
【分類號(hào)】：Q51;TP18
，

本文編號(hào)：1583240