基于GA的入侵檢測分類算法優(yōu)化設(shè)計與實現(xiàn)

2萬字
自己原創(chuàng)的畢業(yè)論文，已經(jīng)通過校內(nèi)系統(tǒng)檢測，重復率低，僅在本站獨家出售，大家放心下載使用

摘 要 計算機網(wǎng)絡(luò)技術(shù)的飛速發(fā)展，使得各種互聯(lián)網(wǎng)絡(luò)新設(shè)備層出不窮，在帶來各種便利的同時，也帶來了越來越普遍和多樣化的各種網(wǎng)絡(luò)安全問題。入侵檢測系統(tǒng)是一種重要的網(wǎng)絡(luò)安全技術(shù)，在眾多的入侵檢測方法中，遺傳算法和神經(jīng)網(wǎng)絡(luò)越來越受到人們的重視，因為遺傳算法和神經(jīng)網(wǎng)絡(luò)都是模擬生物智能化的學習能力而發(fā)展起來的，研究者們通過結(jié)合遺傳算法和神經(jīng)網(wǎng)絡(luò)，并最大限度的利用兩者的優(yōu)點,希望尋找出一種能夠有效解決問題的方案，使得人們更好地理解和學習進化是怎么樣的一個問題。還有運用遺傳算法優(yōu)化神經(jīng)網(wǎng)絡(luò)的最大優(yōu)點是能夠解決神經(jīng)網(wǎng)絡(luò)利用梯度下降法所引起的局部最小值的缺點。
本文介紹了有關(guān)遺傳算法和神經(jīng)網(wǎng)絡(luò)的部分基礎(chǔ)理論，在分析了遺傳算法在神經(jīng)網(wǎng)絡(luò)結(jié)構(gòu)和權(quán)值優(yōu)化應用的基礎(chǔ)上，以Matlab為開發(fā)平臺，利用Matlab神經(jīng)網(wǎng)絡(luò)工具箱和遺傳算法工具箱，運用反向傳播網(wǎng)（BP）算法及遺傳算法實現(xiàn)對KDD99數(shù)據(jù)集的分類分析。即對輸入樣本數(shù)據(jù)進行歸一化處理，創(chuàng)建網(wǎng)絡(luò)模型，進行網(wǎng)絡(luò)學習訓練，通過使用遺傳算法在多個隨機的數(shù)據(jù)中找出最優(yōu)秀的一個或者一組參數(shù)。
最后，將改進的自適應遺傳算法優(yōu)化的神經(jīng)網(wǎng)絡(luò)應用于數(shù)據(jù)預測。對用于預測的樣本數(shù)據(jù)進行標準化處理。將預測結(jié)果與BP算法、遺傳算法優(yōu)化BP網(wǎng)絡(luò)算法的結(jié)果進行了比較，得到較理想的預測效果。

關(guān)鍵詞：遺傳算法；BP算法；入侵檢測；參數(shù)優(yōu)化；分類算法


The classification of intrusion detection
algorithm based on GA to optimize the design and implementation
Abstract The rapid development of computer network technology, make all kinds of Internet new equipment emerge in endlessly, bring convenience in various at the same time, also brought more and more common and diversification of all kinds of network security. Intrusion detection system is a kind of important network security technology, in many of the intrusion detection method, genetic algorithm and neural network is more and more attention by people, Both of Genetic Algorithm and Neutral Network are improving by imitating intelligence of biology．And the research on combining them has become a hot topic at present．People hope to find all efficient way to resolve problems by combining the virtue of both and make full use of them．By the combination people can understand the relationship of learning and evolutionism further．Applying Genetic Algorithm in Neutral Network Can overcome the weakness of running into local minimum which is brought from using Genetic Descent Algorithm．
In the thesis introduces some of the neural network and genetic algorithm of basic theory， the genetic algorithm in neural network structure and the right value optimization on the basis of application, with Matlab for development platform, and use of Matlab neural network and genetic algorithm toolbox, using the back propagation (BP) algorithm and net genetic algorithm for KDD99 set of data classification analysis. That is the sample data input normalized, create network model, the training of network, through the use of genetic algorithms in more random data to find out the best one or a set of parameters.
At the end of the thesis, will improve the adaptive genetic algorithm to optimize the application of neural network to predict。For prediction of the sample data standardization processing。Will predict the results and the BP algorithm, genetic algorithm to optimize the BP neural network algorithm result is compared, get ideal prediction effect.
Key words: Genetic algorithm;BP algorithm;IDS;Parameters optimization;Classification algorithm


目錄
第一章 緒論 1
1.1研究背景與實現(xiàn)意義 1
1.2 研究現(xiàn)狀 1
1.3 研究目的與研究內(nèi)容 3
1.4 論文結(jié)構(gòu)安排 3
第二章 相關(guān)技術(shù)簡介 5
2.1 入侵檢測技術(shù)簡介 5
2.1.1 入侵檢測技術(shù)的定義與分類 5
2.1.2 入侵檢測系統(tǒng)的性能評價指標 5
2.1.3 入侵檢測系統(tǒng)的功能 7
2.2 遺傳算法簡介 7
2.2.1 遺傳算法基本思想 7
2.2.2 遺傳算法的特點 7
2.2.3 遺傳算法的性能指標 8
2.3 開發(fā)平臺簡介 8
2.3.1 Matlab軟件簡介 8
2.3.2 Matlab遺傳算法工具箱簡介 9
第三章 算法設(shè)計與實現(xiàn) 12
3.1 系統(tǒng)結(jié)構(gòu)設(shè)計 12
3.1.1 系統(tǒng)體系結(jié)構(gòu)分類 12
3.1.2 本系統(tǒng)的結(jié)構(gòu)模型 12
3.2 功能模塊的設(shè)計 14
3.2.1 數(shù)據(jù)提取模塊的設(shè)計 14
3.2.2 BP算法模塊的設(shè)計 14
3.2.3 GA-BP算法模塊的設(shè)計 14
3.3 實驗數(shù)據(jù)集的分析 16
3.3.1 分析KDD數(shù)據(jù) 16
3.3.2 KDD CUP99 數(shù)據(jù)集攻擊類型及分布 17
3.4 功能模塊的實現(xiàn) 18
3.4.1 數(shù)據(jù)提取模塊的實現(xiàn) 18
3.4.2 BP算法模塊的實現(xiàn) 20
3.4.3 GA-BP算法模塊的實現(xiàn) 23
3.5 結(jié)果對比分析 27
第四章 總結(jié)與展望 31
4.1 總結(jié) 31
4.2 展望 32
致 謝 32
參考文獻 33