A survey of feature selection in Internet traffic characterization

In the last decade, the research community has focused on new classification methods that rely on statistical characteristics of Internet traffic, instead of pre-viously popular port-number-based or payload-based methods, which are under even bigger constrictions. Some research works based on statistical characteristics generated large fea-ture sets of Internet traffic; however, nowadays it?s impossible to handle hun-dreds of features in big data scenarios, only leading to unacceptable processing time and misleading classification results due to redundant and correlative data. As a consequence, a feature selection procedure is essential in the process of Internet traffic characterization. In this paper a survey of feature selection methods is presented: feature selection frameworks are introduced, and differ-ent categories of methods are briefly explained and compared; several proposals on feature selection in Internet traffic characterization are shown; finally, future application of feature selection to a concrete project is proposed.

A novel information theory method for filter feature selection

In this paper, we propose a novel filter for feature selection. Such filter relies on the estimation of the mutual information between features and classes. We bypass the estimation of the probability density function with the aid of the entropic-graphs approximation of Rényi entropy, and the subsequent approximation of the Shannon one. The complexity of such bypassing process does not depend on the number of dimensions but on the number of patterns/samples, and thus the curse of dimensionality is circumvented. We show that it is then possible to outperform a greedy algorithm based on the maximal relevance and minimal redundancy criterion. We successfully test our method both in the contexts of image classification and microarray data classification.

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.

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%.

On the suitability of Prototype Selection methods for kNN classification with distributed data

In the current Information Age, data production and processing demands are ever increasing. This has motivated the appearance of large-scale distributed information. This phenomenon also applies to Pattern Recognition so that classic and common algorithms, such as the k-Nearest Neighbour, are unable to be used. To improve the efficiency of this classifier, Prototype Selection (PS) strategies can be used. Nevertheless, current PS algorithms were not designed to deal with distributed data, and their performance is therefore unknown under these conditions. This work is devoted to carrying out an experimental study on a simulated framework in which PS strategies can be compared under classical conditions as well as those expected in distributed scenarios. Our results report a general behaviour that is degraded as conditions approach to more realistic scenarios. However, our experiments also show that some methods are able to achieve a fairly similar performance to that of the non-distributed scenario. Thus, although there is a clear need for developing specific PS methodologies and algorithms for tackling these situations, those that reported a higher robustness against such conditions may be good candidates from which to start.

Development and use of a computer simulation model for the evaluation of design and control alternatives for intersections of minor roads with multi-lane rural highways: selection of the simulation model. Interim report.

Federal Highway Administration, Washington, D.C.

German Foreign Ministry Archives 1867-1920 a selection [microform].

For detailed list of contents see typewritten list accompanying the microfilms.

Survey of oceanographic and meteorological parameters of importance to the site selection of an Ocean Food and Energy Farm (OFEF) in the eastern Pacific : final report, July 1975-September 1976 prepared for U.S. Energy Research & Development Administration /

Final report: ERDA, Subtask 5 - OFEF Project.

Decision analysis for abandoned mine reclamation site selection and planning : a report /

"August 1979."

Effects of short term oven aging on volumetrics and selection of N-design /

"Study was conducted in cooperation with the U.S. Department of Transportation, Federal Highway Administration."

Italian wild hop (Humulus lupulus L.): from biodiversity evaluation to varietal selection

Hop (Humulus lupulus L.) is a dioecius perennial plant. The cultivation is specific for female plants, used mainly for brewing and pharmacology. Female inflorescence, known as cone or strobili, contains bitter acids, essential oil and polyphenols. Commercial hop cultivation provides better results in regions between 45 and 55 degrees north or south in latitude, an area that also includes the northern part of Italy, where hop is endemic. Despite several studies have been conducted on the characterization of wild hops biodiversity in the U.S.A. and Europe, a lack in literature concerning the description of Italian wild hops genetic variability is still present. The increasing request of hop varieties improved in important traits, like diseases, resistance and valuable aroma profile, is bringing the hop industry. Moreover, Italian agricultural sector needs new impulse to be competitive in the market. In this view, Italian wild hop biodiversity is a resource, useful for the obtaining of Italian hop varieties, characterized by peculiar aromatic traits and more adaptable to Mediterranean climate, making their cultivation more sustainable. Based on this consideration, the present Ph.D. thesis deals with the evaluation of the Italian hop biodiversity, through the characterization of the wild samples under different point of view. The project started with the recovery of wild hop samples in different areas of north of Italy to consitue a collection field, where 11 commercial cultivars of US and European origin were grown, to have a complete vision of the hop panorama. Ph.D. project followed different research lines, the results of each one contributed to completly characterize the northern Italian hop wild biodiversity: • the morphological description showed a high phenological variability (Study 1); • the genetic characterization confirmed the rich biodiversity of the Italian population and showed a significant genetic distance between Italian genotypes and the commercial cultivars, taken in consideration (Study 2); • the need of an early sex discrimination method leads to an improvement of a genetic marker, developing a more efficient marker (Study 3); • a complete morphologic, genetic and chemical analysis of plants gave results to select the most promising genotypes (Study 4); • the comparison between the performance of wild hops and commercial cultivars in the same collection field indicated that some wild genotypes had a higher environment adaptability (Study 5); • the evaluation of the terroir, obtained comparing commercial cultivars in the collection field and the same genotypes purchased in the market, showed the influence of the northern Italian environment on the aromatic profile (Study 5); • a new analytical method for the revelation of bioactive metabolites and a simple extraction procedure were developed (Study 6). In conclusion, the Ph.D. thesis, contains the first characterization of Italian wild hop, made under field condition. The present study: i) permits to obtain a complete and significative description of the genotypes; ii) allows the identification of the most promising wild Italian genotypes; iii) allows the identification of commercial cultivars more adaptable the northern Italian climate.

Expert systems for vendor selection

The survival of organisations, especially SMEs, depends, to the greatest extent, on those who supply them with the required material input. This is because if the supplier fails to deliver the right materials at the right time and place, and at the right price, then the recipient organisation is bound to fail in its obligations to satisfy the needs of its customers, and to stay in business. Hence, the task of choosing a supplier(s) from a list of vendors, that an organisation will trust with its very existence, is not an easy one. This project investigated how purchasing personnel in organisations solve the problem of vendor selection. The investigation went further to ascertain whether an Expert Systems model could be developed and used as a plausible solution to the problem. An extensive literature review indicated that very scanty research has been conducted in the area of Expert Systems for Vendor Selection, whereas many research theories in expert systems and in purchasing and supply management chain, respectively, had been reported. A survey questionnaire was designed and circulated to people in the industries who actually perform the vendor selection tasks. Analysis of the collected data confirmed the various factors which are considered during the selection process, and established the order in which those factors are ranked. Five of the factors, namely, Production Methods Used, Vendors Financial Background, Manufacturing Capacity, Size of Vendor Organisations, and Suppliers Position in the Industry; appeared to have similar patterns in the way organisations ranked them. These patterns suggested that the bigger the organisation, the more importantly they regarded the above factors. Further investigations revealed that respondents agreed that the most important factors were: Product Quality, Product Price and Delivery Date. The most apparent pattern was observed for the Vendors Financial Background. This generated curiosity which led to the design and development of a prototype expert system for assessing the financial profile of a potential supplier(s). This prototype was called ESfNS. It determines whether a prospective supplier(s) has good financial background or not. ESNS was tested by the potential users who then confirmed that expert systems have great prospects and commercial viability in the domain for solving vendor selection problems.

Selection and application of risk management tools and techniques for build-operate-transfer projects

Construction projects are risky. However, the characteristics of the risk highly depend on the type of procurement being adopted for managing the project. A build-operate-transfer (BOT) project is recognized as one of the most risky project schemes. There are instances of project failure where a BOT scheme was employed. Ineffective rts are increasingly being managed using various risk management tools and techniques. However, application of those tools depends on the nature of the project, organization's policy, project management strategy, risk attitude of the project team members, and availability of the resources. Understanding of the contents and contexts of BOT projects, together with a thorough understanding of risk management tools and techniques, helps select processes of risk management for effective project implementation in a BOT scheme. This paper studies application of risk management tools and techniques in BOT projects through reviews of relevant literatures and develops a model for selecting risk management process for BOT projects. The application to BOT projects is considered from the viewpoints of the major project participants. Discussion is also made with regard to political risks. This study would contribute to the establishment of a framework for systematic risk management in BOT projects.

Practice report: the role of environmental factors in industrial site selection activities: a case of limestone quarry expansion in Barbados, West Indies

Site selection is a key activity for quarry expansion to support cement production, and is governed by factors such as resource availability, logistics, costs, and socio-environmental factors. Adequate consideration of all factors facilitates both industrial productivity and sustainable economic growth. This study illustrates the site selection process that was undertaken for the expansion of limestone quarry operations to support cement production in Barbados. First, alternate sites with adequate resources to support a 25-year development horizon were identified. Second, socio-environmental conditions were described and potential impacts identified. Third, a comparative matrix was constructed to evaluate relative site characteristics with respect to physical, ecological, socio-cultural and economic factors. The study shows that environmental factors were essential to the final site recommendation.

Development of an e-business selection framework for manufacturing SMEs : a study of the printing industry

The aim of the research is to develop an e-business selection framework for small and medium enterprises (SMEs) by integrating established techniques in planning. The research is case based, comprising four case studies carried out in the printing industry for the purpose of evaluating the framework. Two of the companies are from Singapore, while the other two are from Guangzhou, China and Jinan, China respectively. To determine the need of an e-business selection framework for SMEs, extensive literature reviews were carried out in the area of e-business, business planning frameworks, SMEs and the printing industry. An e-business selection framework is then proposed by integrating the three established techniques of the Balanced Scorecard (BSC), Value Chain Analysis (VCA) and Quality Function Deployment (QFD). The newly developed selection framework is pilot tested using a published case study before actual evaluation is carried out in four case study companies. The case study methodology was chosen because of its ability to integrate diverse data collection techniques required to generate the BSC, VCA and QFD for the selection framework. The findings of the case studies revealed that the three techniques of BSC, VCA and QFD can be integrated seamlessly to complement on each other’s strengths in e-business planning. The eight-step methodology of the selection framework can provide SMEs with a step-by-step approach to e-business through structured planning. Also, the project has also provided better understanding and deeper insights into SMEs in the printing industry.