American scientific publishing group

With the help of Internet, the online news can be instantly spread around the world. Most of peoples now have the habit of reading and sharing news online, for instance, using social media like Twitter and Facebook. Typically, the news popularity can be indicated by the number of reads, likes or shares. For the online news stake holders such as content providers or advertisers, it’s very valuable if the popularity of the news articles can be accurately predicted prior to the publication. Thus, it is interesting and meaningful to use the machine learning techniques to predict the popularity of online news articles. Various works have been done in prediction of online news popularity. Popularity of news depends upon various features like sharing of online news on social media, comments of visitors for news, likes for news articles etc. It is necessary to know what makes one online news article more popular than another article. Unpopular articles need to get optimize for further popularity. In this paper, different methodologies are analyzed which predict the popularity of online news articles. These methodologies are compared, their parameters are considered and improvements are suggested. The proposed methodology describes online news popularity predicting system.

Heart disease is a leading cause of mortality worldwide, and early detection and accurate prediction of heart disease can significantly improve patient outcomes. Machine learning techniques have shown great promise in assisting healthcare professionals in diagnosing and predicting heart disease. The diagnosis and prognosis of heart disease must be improved, refined, and accurate, because a small mistake can cause weakness or death. According to a recent World Health Organization study, 17.5 million people die each year. By 2030, this number will increase to 75 million.[2] This document explains how to enable online KSRM capabilities. The KSRM smart system allows users to report heart-related problems. This research paper aims to explore the use of machine learning algorithms for effective heart disease prediction classification with Ada boost for improve the accuracy of algorithm.

In Mobile Ad hoc Network (MANET) all the nodes are freely moves in the absence of without ant centralized coordination system. Due to that the attackers or malicious nodes are easily affected that kind of network and responsible for the routing misbehavior. The routing is network is mandatory to deliver data in between source and destination. In this research we work on security field in MANET and proposed a novel security scheme against routing misbehavior through Black hole attack. The Ad hoc On demand Multipath Routing (AOMDV) protocol is consider for routing and also to improves the routing quality as compare to single path routing protocol. The attacker is affected all the possible paths that is selected by sender for sending data in network. The malicious nodes are forward optimistic reply at the time of routing by that their identification is also a complex procedure. The proposed Intrusion Detection System (IDS) scheme is identified the attacker information through hop count mechanism. The routing information of actual data is reached to which intermediate node and the next hop information is exist at that node is confirm by IDS scheme. The black hole attacker node Identification (ID) is forward in network by that in future attacker is not participating in routing procedure. The proposed security scheme detects and provides the deterrence against routing misbehavior through malicious attack. Here we compare the routing performance of AOMDV, Attack and IDS scheme. The performance of normal multipath routing and proposed IDS scheme is almost equal. The attacker has degrades the whole routing performance but observed that in presence of attacker, routing misbehavior is completely block by the proposed IDS scheme and recovers 95 % of data as compare to normal routing. Keywords- -AOMDV, MANET, IDS, Black hole attack, Routing misbehavior.