Feature selection by multi-objective optimisation: Application to network anomaly detection by hierarchical self-organising maps

Feature selection is an important and active issue in clustering and classification problems. By choosing an adequate feature subset, a dataset dimensionality reduction is allowed, thus contributing to decreasing the classification computational complexity, and to improving the classifier performance by avoiding redundant or irrelevant features. Although feature selection can be formally defined as an optimisation problem with only one objective, that is, the classification accuracy obtained by using the selected feature subset, in recent years, some multi-objective approaches to this problem have been proposed. These either select features that not only improve the classification accuracy, but also the generalisation capability in case of supervised classifiers, or counterbalance the bias toward lower or higher numbers of features that present some methods used to validate the clustering/classification in case of unsupervised classifiers. The main contribution of this paper is a multi-objective approach for feature selection and its application to an unsupervised clustering procedure based on Growing Hierarchical Self-Organising Maps (GHSOMs) that includes a new method for unit labelling and efficient determination of the winning unit. In the network anomaly detection problem here considered, this multi-objective approach makes it possible not only to differentiate between normal and anomalous traffic but also among different anomalies. The efficiency of our proposals has been evaluated by using the well-known DARPA/NSL-KDD datasets that contain extracted features and labelled attacks from around 2 million connections. The selected feature sets computed in our experiments provide detection rates up to 99.8% with normal traffic and up to 99.6% with anomalous traffic, as well as accuracy values up to 99.12%.

Assessment of a simplified experimental procedure to evaluate impact sound reduction of floor coverings

The impact noise reduction provided by floor coverings is usually obtained in laboratory, using the methodology described in the standard EN ISO 140-8, which requires the use of standard acoustic chambers. The construction of such chambers, following the requirements described in the EN ISO 140-1, implies a significant investment, and therefore only a limited number exists in each country. Alternatives to these standard methodologies, that allow a sufficiently accurate evaluation and require lower resources, have been interesting many researchers and manufacturers. In this paper, one such strategy is discussed, where a reduced sized slab is used to determine the noise reduction provided by floor coverings, following the procedure described in the ISO/CD 16251-1 technical document. Several resilient coverings, floating floors and floating slabs are tested and the results are compared with those obtained using the procedures described in the standards EN ISO 140-8 and EN ISO 717-2.

Improving kNN multi-label classification in Prototype Selection scenarios using class proposals

Prototype Selection (PS) algorithms allow a faster Nearest Neighbor classification by keeping only the most profitable prototypes of the training set. In turn, these schemes typically lower the performance accuracy. In this work a new strategy for multi-label classifications tasks is proposed to solve this accuracy drop without the need of using all the training set. For that, given a new instance, the PS algorithm is used as a fast recommender system which retrieves the most likely classes. Then, the actual classification is performed only considering the prototypes from the initial training set belonging to the suggested classes. Results show that this strategy provides a large set of trade-off solutions which fills the gap between PS-based classification efficiency and conventional kNN accuracy. Furthermore, this scheme is not only able to, at best, reach the performance of conventional kNN with barely a third of distances computed, but it does also outperform the latter in noisy scenarios, proving to be a much more robust approach.

Performance Assessment of a Polymer Electrolyte Membrane Electrochemical Reactor under Alkaline Conditions – A Case Study with the Electrooxidation of Alcohols

A novel polymer electrolyte membrane electrochemical reactor (PEMER) configuration has been employed for the direct electrooxidation of propargyl alcohol (PGA), a model primary alcohol, towards its carboxylic acid derivatives in alkaline medium. The PEMER configuration comprised of an anode and cathode based on nanoparticulate Ni and Pt electrocatalysts, respectively, supported on carbonaceous substrates. The electrooxidation of PGA was performed in 1.0 M NaOH, where a cathode based on a gas diffusion electrode was manufactured for the reduction of oxygen in alkaline conditions. The performance of a novel alkaline anion-exchange membrane based on Chitosan (CS) and Poly(vinyl) alcohol (PVA) in a 50:50 composition ratio doped with a 5 wt.% of poly (4-vinylpyridine) organic ionomer cross-linked, methyl chloride quaternary salt resin (4VP) was assessed as solid polymer electrolyte. The influence of 4VP anionic ionomer loading of 7, 12 and 20 wt.% incorporated into the electrocatalytic layers was examined by SEM and cyclic voltammetry (CV) upon the optimisation of the electroactive area, the mechanical stability and cohesion of the catalytic ink onto the carbonaceous substrate for both electrodes. The performance of the 4VP/CS:PVA membrane was compared with the commercial alkaline anion-exchange membrane FAA −a membrane generally used in direct alcohol alkaline fuel cells- in terms of polarisation plots in alkaline conditions. Furthermore, preparative electrolyses of the electrooxidation of PGA was performed under alkaline conditions of 1 M NaOH at constant current density of 20 mA cm−2 using a PEMER configuration to provide proof of the principle of the feasibility of the electrooxidation of other alcohols in alkaline media. PGA conversion to Z isomers of 3-(2-propynoxy)-2-propenoic acid (Z-PPA) was circa 0.77, with average current efficiency of 0.32. Alkaline stability of the membranes within the PEMER configuration was finally evaluated after the electrooxidation of PGA.

Application-Based Analysis of Register File Criticality for Reliability Assessment in Embedded Microprocessors

There is an increasing concern to reduce the cost and overheads during the development of reliable systems. Selective protection of most critical parts of the systems represents a viable solution to obtain a high level of reliability at a fraction of the cost. In particular to design a selective fault mitigation strategy for processor-based systems, it is mandatory to identify and prioritize the most vulnerable registers in the register file as best candidates to be protected (hardened). This paper presents an application-based metric to estimate the criticality of each register from the microprocessor register file in microprocessor-based systems. The proposed metric relies on the combination of three different criteria based on common features of executed applications. The applicability and accuracy of our proposal have been evaluated in a set of applications running in different microprocessors. Results show a significant improvement in accuracy compared to previous approaches and regardless of the underlying architecture.