This dissertation proposes a novel mathematical model for the Amoeba or the Nelder-Mead simplex optimization (NM) algorithm. The proposed Hassan NM (HNM) algorithm allows components of the reflected vertex to adapt to different operations, by breaking down the complex structure of the simplex into multiple triangular simplexes that work sequentially to optimize the individual components of mathematical functions. When the next formed simplex is characterized by different operations, it gives the simplex similar reflections to that of the NM algorithm, but with rotation through an angle determined by the collection of nonisometric features. As a consequence, the generating sequence of triangular simplexes is guaranteed that not only they have different shapes, but also they have different directions, to search the complex landscape of mathematical problems and to perform better performance than the traditional hyperplanes simplex. To test reliability, efficiency, and robustness, the proposed algorithm is examined on three areas of large-scale optimization categories: systems of nonlinear equations, nonlinear least squares, and unconstrained minimization. The experimental results confirmed that the new algorithm delivered better performance than the traditional NM algorithm, represented by a famous Matlab function, known as fminsearch.
In addition, the new trigonometric simplex design provides a platform for further development of reliable and robust sparse autoencoder software (SAE) for intrusion detection system (IDS) applications. The proposed error function for the SAE is designed to make a trade-off between the latent state representation for more mature features and network regularization by applying the sparsity constraint in the output layer of the proposed SAE network. In addition, the hyperparameters of the SAE are tuned based on the HNM algorithm and were proved to give a better capability of extracting features in comparison with the existing developed algorithms. In fact, the proposed SAE can be used for not only network intrusion detection systems, but also other applications pertaining to deep learning, feature extraction, and pattern analysis. Results from experimental tests showed that the different layers of the enhanced SAE could efficiently adapt to various levels of learning hierarchy. Finally, additional tests demonstrated that the proposed IDS architecture could provide a more compact and effective immunity system for different types of network attacks with a significant detection accuracy of 99.63% and an F-measure of 0.996, on average, when penalizing sparsity constraint directly on the synaptic weights within the network.
|Commitee:||Patra, Prabir , Dichter, Julius , Faezipour, Miad , Tokgoz, Emre|
|School:||University of Bridgeport|
|Department:||Computer Science and Engineering|
|School Location:||United States -- Connecticut|
|Source:||DAI-A 82/5(E), Dissertation Abstracts International|
|Subjects:||Computer Engineering, Design, Applied Mathematics|
|Keywords:||Advanced sparse autoencoder, Hyperparameter optimization, Intrusion detection system, Systems of nonlinear equations, Trigonometric simplex designs, Unconstrained minimization, fminsearch|
Copyright in each Dissertation and Thesis is retained by the author. All Rights Reserved
The supplemental file or files you are about to download were provided to ProQuest by the author as part of a
dissertation or thesis. The supplemental files are provided "AS IS" without warranty. ProQuest is not responsible for the
content, format or impact on the supplemental file(s) on our system. in some cases, the file type may be unknown or
may be a .exe file. We recommend caution as you open such files.
Copyright of the original materials contained in the supplemental file is retained by the author and your access to the
supplemental files is subject to the ProQuest Terms and Conditions of use.
Depending on the size of the file(s) you are downloading, the system may take some time to download them. Please be