Multi-source data analysis and evaluation of machine learning techniques for SQL injection detection

Security in edge-assisted Internet of Things: challenges and solutions

“…In this subsection, we conduct extensive simulations to illustrate how to adjust the complexity of security schemes by using the proposed solution. Two existing security schemes against DDoS attack [18] and malware injection attack [24] are applied in this case study. We present the performance of the two schemes before and after feature selection to validate the effectiveness of our proposed solution to address security-energy tradeoff.…”

“…Through the feature selection performed by the proposed solution, different tradeoffs between security and energy can be reached depending on the demands of IoT applications. After feature selection, the state-of-art security schemes [18,24] are applied against DDoS attack and malware injection attack, respectively. Figure 6 presents performances of two schemes with different numbers of selected features, where the balanced accuracy reflects the security level and computational time reflects the energy consumption.…”

“…For instance, NLP is applied to identify structured query language (SQL) injection vulnerabilities in edge node program [23]. Multiple other machine learning techniques are also evaluated by Ross et al [24] to compare the accuracy in detecting SQL injection attacks. A cross-site scripting (XSS) classifier [25] is proposed based on three features of URLs, webpage, and social networking services to detect XSS vulnerabilities, achieving over 97% accuracy.…”

Construction of a parametric model of competitive access in relational databases by using a random forest method

“…The library Accord.NET is often used in research into the application of machine learning algorithms [15,16], including the studies into similar topics [17,19]. The closest analogs to this library are the commercially-available suite MATLAB, the R language implementation, as well as the free library scikit-learn.…”

“…Consequently, without additional transformations, the logically interrelated attributes would be considered independent of each other. Such a simplification would reduce the generalizing capacity of the algorithm, as well as increase the costs of keeping a forest and the complexity of its construction [17]. To resolve this task, the modifications to the original method are proposed.…”