现代电子技术
現代電子技術
현대전자기술
Modern Electronics Technique
2015年
23期
85-88
,共4页
网络安全%BP神经网络%人工鱼群算法%自适应算法
網絡安全%BP神經網絡%人工魚群算法%自適應算法
망락안전%BP신경망락%인공어군산법%자괄응산법
network security%BP neural network%artificial fish swarm algorithm%adaptive algorithm
针对现代计算机网络存在安全评价指标种类多、各指标的非线性特性强等，采用传统评价方法评估存在操作难度大、精确性低等问题，提出了基于自适应BP神经网络的计算机网络安全评价算法。算法采用建立人工鱼群优化算法对传统BP神经网络结构参数进行优化确定，然后根据优化后的网络对样本指标进行训练和学习，最后再通过专家评估测试数据对算法的有效性进行检测与评估。仿真试验结果验证了算法在计算机网络安全评价方面的有效性。
針對現代計算機網絡存在安全評價指標種類多、各指標的非線性特性彊等，採用傳統評價方法評估存在操作難度大、精確性低等問題，提齣瞭基于自適應BP神經網絡的計算機網絡安全評價算法。算法採用建立人工魚群優化算法對傳統BP神經網絡結構參數進行優化確定，然後根據優化後的網絡對樣本指標進行訓練和學習，最後再通過專傢評估測試數據對算法的有效性進行檢測與評估。倣真試驗結果驗證瞭算法在計算機網絡安全評價方麵的有效性。
침대현대계산궤망락존재안전평개지표충류다、각지표적비선성특성강등，채용전통평개방법평고존재조작난도대、정학성저등문제，제출료기우자괄응BP신경망락적계산궤망락안전평개산법。산법채용건립인공어군우화산법대전통BP신경망락결구삼수진행우화학정，연후근거우화후적망락대양본지표진행훈련화학습，최후재통과전가평고측시수거대산법적유효성진행검측여평고。방진시험결과험증료산법재계산궤망락안전평개방면적유효성。
Since the modern computer network has the problems of various security evaluation indicators and strong nonlinear characteristic of each indicator,and the traditional evaluation method has difficult operation and low accuracy,an evaluation al?gorithm of computer network security based on adaptive BP neural network is proposed. In the algorithm,the artificial fish swarm algorithm(AFSA)is established to optimize and determine the structure parameters of the traditional BP neural network, and then the sample indicators are trained and learned according to the optimized network,finally the validity of the algorithm is tested and evaluated by means of the test data of expert evaluation. The simulation experiment results show that the algorithm has the validity in the evaluation of computer network security.
