A Predictive Classification Method for Email Phishing Attacks u sing Random Forest and A R trees

Abstract

Concrete and steel structures influences the construction of multi-storey structures. The aid of progressive collapse increases when there is a failure of one or more load bearing structural elements. Thereafter, this case study is carried out to determine the prospective of the progressive collapse of an irregular (L shaped) building due to the failure or removal of two adjacent columns at a time present in the ground floor. Failure may occur because of the natural or manmade accidental loads like explosion or seismic loads, collision of vehicles, etc. Columns at different locations were removed and the slab loads had been increased as per the General Services Administration (GSA) guidelines and the results in terms of Demand Capacity Ratios (DCR) are compared for all the cases. The Demand to Capacity Ratios were calculated for the interested columns. It is observed that when the interior columns were removed then the possibility of progressive collapse is more. This study has been made for the case or earthquake forces for corresponding zone II and zone V.Cyber-attacks are the attempts made by an individual or an organization deliberately, to breach the information system mainly computers of another individual or organization. These attacks have risen in recent years due to various reasons posing the need for systems that can use adaptive learning techniques to detect and mitigate these attacks at an early stage. Phishing is one of the significant cyber-attacks. According to global security report 2019, phishing was the major cause of attacks in corporate networks. Phishing attack uses disguised email to achieve its goal. In this attack, attacker masquerade himself as a trusted individual or a company and trick the email recipient into clicking malicious links or attachments. The proposed method provides a testbed for detecting and mitigating various types of phishing attacks. Machine learning techniques are used to build an intelligent system which can detect phishing attacks. This application uses random forest algorithm with AR-Trees (acceptance-rejection tree algorithm) to determine the attacks by considering various datasets available online and new datasets dynamically constructed for making the system ready to mitigate future phishing attacks.