As novel technologies continue to reshape the digital era, cyberattacks are also increasingly becoming more commonplace and sophisticated. Distributed denial of service (DDoS) attacks are, perhaps, the most prevalent and exponentially-growing attack, targeting the varied and emerging computational network infrastructures across the globe. This necessitates the design of an efficient and early detection of large-scale sophisticated DDoS attacks. Software defined networks (SDN) point to a promising solution, as a network paradigm which decouples the centralized control intelligence from the forwarding logic. In this work, a deep convolutional neural network (CNN) ensemble framework for efficient DDoS attack detection in SDNs is proposed. The proposed framework is evaluated on a current state-of-the-art Flow-based dataset under established benchmarks. Improved accuracy is demonstrated against existing related detection approaches. INDEX TERMS Software defined network (SDN), anomaly detection, distributed denial of service (DDoS), deep learning, deep convolutional neural network (CNN).
Software organization always aims at developing a quality software product using the estimated development resources, effort, and time. Global Software Development (GSD) has emerged as an essential tool to ensure optimal utilization of resources, which is performed in globally distributed settings in various geographical locations. Global software engineering focuses on reducing the cost, increasing the development speed, and accessing skilled developers worldwide. Estimating the required amount of resources and effort in the distributed development environment remains a challenging task. Thus, there is a need to focus on cost estimation models in the GSD context. We nevertheless acknowledge that several cost estimation techniques have been reported. However, to the best of our knowledge, the existing cost estimation techniques/models lack considering the additional cost drivers required to compute the accurate cost estimation in the GSD context. Motivated by this, the current work aims at identifying the other cost drivers that affect the cost estimation in the context of GSD. To achieve the targeted objectives, current state-of-the-art related to existing cost estimation techniques of GSD is reported. We adopted SLR and Empirical approach to address the formulated research questions. The current study also identifies the missing factors that would help the practitioners improve the cost estimation models. The results indicate that previously conducted work ignores the additional elements necessary for the cost estimation in the GSD context. Moreover, the current work proposes a conceptual cost estimation model tailored to fit the GSD context.
An activity cliff (AC) is formed by a pair of structurally similar compounds with a large difference in potency. Accordingly, ACs reveal structure–activity relationship (SAR) discontinuity and provide SAR information for compound optimization. Herein, we have investigated the question if ACs could be predicted from image data. Therefore, pairs of structural analogs were extracted from different compound activity classes that formed or did not form ACs. From these compound pairs, consistently formatted images were generated. Image sets were used to train and test convolutional neural network (CNN) models to systematically distinguish between ACs and non-ACs. The CNN models were found to predict ACs with overall high accuracy, as assessed using alternative performance measures, hence establishing proof-of-principle. Moreover, gradient weights from convolutional layers were mapped to test compounds and identified characteristic structural features that contributed to successful predictions. Weight-based feature visualization revealed the ability of CNN models to learn chemistry from images at a high level of resolution and aided in the interpretation of model decisions with intrinsic black box character.
Activity landscapes (ALs) are graphical representations that combine compound similarity and activity data. ALs are constructed for visualizing local and global structure–activity relationships (SARs) contained in compound data sets. Three-dimensional (3D) ALs are reminiscent of geographical maps where differences in landscape topology mirror different SAR characteristics. 3D AL models can be stored as differently formatted images and are thus amenable to image analysis approaches, which have thus far not been considered in the context of graphical SAR analysis. In this proof-of-concept study, 3D ALs were constructed for a variety of compound activity classes and 3D AL image variants of varying topology and information content were generated and classified. To these ends, convolutional neural networks (CNNs) were initially applied to images of original 3D AL models with color-coding reflecting compound potency information that were taken from different viewpoints. Images of 3D AL models were transformed into variants from which one-dimensional features were extracted. Other machine learning approaches including support vector machine (SVM) and random forest (RF) algorithms were applied to derive models on the basis of such features. In addition, SVM and RF models were trained using other features obtained from images through edge filtering. Machine learning was able to accurately distinguish between 3D AL image variants with different topology and information content. Overall, CNNs which directly learned feature representations from 3D AL images achieved highest classification accuracy. Predictive performance for CNN, SVM, and RF models was highest for image variants emphasizing topological elevation. In addition, SVM models trained on rudimentary images from edge filtering classified such images with high accuracy, which further supported the critical role of altitude-dependent topological features for image analysis and predictions. Taken together, the findings of our proof-of-concept investigation indicate that image analysis has considerable potential for graphical SAR exploration to systematically infer different SAR characteristics from topological features of 3D ALs.
The goal of software process improvement (SPI) is to improve software processes and produce high-quality software, but the results of SPI efforts in small-and medium-sized enterprises (SMEs) that develop software have been unsatisfactory. The objective of this study is to support the prolific and successful CMMI-based implementation of SPI in SMEs by presenting the facts related to the unofficial adoption of CMMI level 2 process area-specific practices by software SMEs. Two questionnaire surveys were performed, and 42 questionnaires were selected for data analysis. The questionnaires were filled out by experts from 42 non-CMMI-certified software SMEs based in Malaysia and Pakistan. In the case of each process area of CMMI level 2, the respondents were asked to choose from three categories, namely 'below 50 %,' '50-75 %,' and 'above 75 %'. The percentages indicated the extent to which process area-specific practices are routinely followed in the respondents' respective organizations. To deal with differing standards for defining SMEs, the notion of the common range standard has been introduced. The results of the study show that a large segment of software development SMEs informally follows the specific practices of CMMI level 2 process areas and thus has true potential for rapid and effective CMMI-based SPI. The results further indicate that, in the case of four process areas of CMMI level 2, there are statistically significant differences between the readiness of small and medium software enterprises to adopt the specific practices of those process areas, and between trends on their part to do so unofficially. The findings, manifesting various degrees of unofficial readiness for CMMI-based SPI among SMEs, can be used to define criteria for the selection of SMEs that would be included in SPI initiatives funded by relevant authorities. In the interests of developing fruitful CMMI-based SPI and to enhance the success rate of CMMI-based SPI initiatives, the study suggests that 'ready' or 'potential' SMEs should be given priority for SPI initiatives.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
334 Leonard St
Brooklyn, NY 11211
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.