Iffat A. Gheyas

Feature subset selection in large dimensionality domains

Searching for an optimal feature subset from a high dimensional feature space is known to be an NP-complete problem. We present a hybrid algorithm, SAGA, for this task. SAGA combines the ability to avoid being trapped in a local minimum of simulated annealing with the very high rate of convergence of the crossover operator of genetic algorithms, the strong local search ability of greedy algorithms and the high computational efficiency of generalized regression neural networks. We compare the performance over time of SAGA and well-known algorithms on synthetic and real datasets. The results show that SAGA outperforms existing algorithms.

A neural network-based framework for the reconstruction of incomplete data sets

The treatment of incomplete data is an important step in the pre-processing of data. We propose a novel nonparametric algorithm Generalized regression neural network Ensemble for Multiple Imputation (GEMI). We also developed a single imputation (SI) version of this approach-GESI. We compare our algorithms with 25 popular missing data imputation algorithms on 98 real-world and synthetic datasets for various percentage of missing values. The effectiveness of the algorithms is evaluated in terms of (i) the accuracy of output classification: three classifiers (a generalized regression neural network, a multilayer perceptron and a logistic regression technique) are separately trained and tested on the dataset imputed with each imputation algorithm, (ii) interval analysis with missing observations and (iii) point estimation accuracy of the missing value imputation. GEMI outperformed GESI and all the conventional imputation algorithms in terms of all three criteria considered.

A novel neural network ensemble architecture for time series forecasting

We propose a novel homogeneous neural network ensemble approach called Generalized Regression Neural Network (GEFTS-GRNN) Ensemble for Forecasting Time Series, which is a concatenation of existing machine learning algorithms. GEFTS uses a dynamic nonlinear weighting system wherein the outputs from several base-level GRNNs are combined using a combiner GRNN to produce the final output. We compare GEFTS with the 11 most used algorithms on 30 real datasets. The proposed algorithm appears to be more powerful than existing ones. Unlike conventional algorithms, GEFTS is effective in forecasting time series with seasonal patterns.

Contagion in cyber security attacks

Systems security is essential for the efficient operation of all organizations. Indeed, most large firms employ a designated ‘Chief Information Security Officer’ to coordinate the operational aspects of the organization’s information security. Part of this role is in planning investment responses to information security threats against the firm’s corporate network infrastructure. To this end, we develop and estimate a vector equation system of threats to 10 important IP services, using industry standard SANS data on threats to various components of a firm’s information system over the period January 2003 – February 2011. Our results reveal strong evidence of contagion between such attacks, with attacks on ssh and Secure Web Server indicating increased attack activity on other ports. Security managers who ignore such contagious inter-relationships may underestimate the underlying risk to their systems’ defence of security attributes, such as sensitivity and criticality, and thus delay appropriate information security investments.

Reconstruction of Cross-Sectional Missing Data Using Neural Networks

The treatment of incomplete data is an important step in the pre-processing of data. We propose a non-parametric multiple imputation algorithm (GMI) for the reconstruction of missing data, based on Generalized Regression Neural Networks (GRNN). We compare GMI with popular missing data imputation algorithms: EM (Expectation Maximization) MI (Multiple Imputation), MCMC (Markov Chain Monte Carlo) MI, and hot deck MI. A separate GRNN classifier is trained and tested on the dataset imputed with each imputation algorithm. The imputation algorithms are evaluated based on the accuracy of the GRNN classifier after the imputation process. We show the effectiveness of our proposed algorithm on twenty-six real datasets.

Resilience in Information Stewardship

Information security is concerned with protecting the confi- dentiality, integrity, and availability of information systems. System managers deploy their resources with the aim of maintaining target levels of these attributes in the presence of reactive threats. Information stewardship is the challenge of maintaining the sustainability and resilience of the security attributes of (complex, interconnected, multi-agent) information ecosystems. In this paper, we present, in the tradition public economics, a model of stewardship which addresses directly the question of resilience. We model attacker-target-steward behaviour in a fully endogenous Nash equilibrium setting. We analyse the occurrence of externalities across targets and assess the steward’s ability to internalize these externalities under varying informational assumptions. We apply and simulate this model in the case of a critical national infrastructure example.