This paper is to improve the privacy and security in the distributed virtual environment using blockchain technology. One of the feature it provides is greater security in the decentralized virtual environment. A key aspect of this technology is used for various fields like healthcare, finance, business and cloud environment. Key issue of the virtual environment is to protect the data privacy and security which is difficult to handle. To overcome this issue, a new security model to protect the virtual environment is created and will focus on different types of attacks in blockchain technology.

Blockchain security in virtual environment

On the lap of this present-day epoch, Human Activity Recognition (HAR) has been of considerable assistance in case of health monitoring and recovery. The utilization of machine learning approach integration with intelligent agent in the area of health informatics collected via Human Activity Recognition enhances the decision making quality and significance. Its specific homogenization into the Smart Healthcare Monitoring permits gathering, examining and learning from Internet of Things (IoT) wearable devices, undoubtedly achieving knowledge and making analysis on the patient’s state. Despite several research works conducted on Smart Healthcare Monitoring, there remains certain amount of pitfalls, like, time, overhead involved in analysis and also the falsification of analysis. To focus on these issues, a Statistical Partial Regression and Support Vector Intelligent Agent Learning (SPR-SVIAL) for Smart Healthcare Monitoring is proposed. The SPR-SVIAL method is split into two stages. First, data pre-processing along with the dimensionality reduced features are extracted by employing Statistical Partial Regression Feature Extraction model. Here, with the ceaseless thump to-pulsate heart information, triaxial accelerometer information, rest quality, actual work and mental attributes obtained from the input dataset acquired from IoT wearable devices, Partial Least Square is applied to extract the dimensionality reduced features, therefore contributing to Smart Healthcare Monitoring time and accuracy. Next, with these resultant features, Support Vector Intelligent Agent Learning is proposed for Smart Healthcare Monitoring that with the aid of Machine Learning and Intelligent Agent not only reduces the falsification of analysis but also reduces the overhead incurred. The SPR-SVIAL method is tested on simulators and the obtained results indicated better performance upon comparison with the other methods. The results show that we can reduce the time, overhead, false positive rate for healthcare monitoring and achieve a high accuracy rate by performing feature extraction for each of the data recording.

Intelligent cloud agent based action recognition detection using machine learning

Fraud is one of the major problems around the world today. Example: Stock market. In this paper, circular trading is
detected using graphs. We detect the fraud using graph and machine learning. To resolve this problem, different fraud detection
algorithm and approaches were used. In our proposed system, we are using an algorithm called collusion clustering, which is
particularly designed for finding fraudulent sets. This algorithm will detect the fraud more efficiently convey the result.

Fraud Detection in Stock Market Using Collusion Clustering Algorithm

Heart sickness is a main source for a lot of passings around the globe, and it is fundamental for each person to take great consideration of their heart. Cardiovascular infections (CVDs) turned into a significant reason for casualty in India. More than 54.5 million individuals experiencing CVDs in 2016 and one out of 4 passings are presently because of coronary illness. This task adds to conveying the expectation model for a coronary illness or respiratory failure utilizing AI and to make an easy to understand portable application that can get the information from the client’s savvy band or smartwatch for prompt forecasts. This tells individuals their coronary illness well and takes adequate measures to forestall heart absconds ahead of time. An alternate organic and actual boundaries like age, sex, pulse, circulatory strain, cholesterol level and chest torment locale can be utilized to estimate. The preparation and examination dataset, which comprises 14 distinct ascribes, was downloaded from kaggle.com. Managed calculations in AI will in general focus on information and the outcomes show that the proficiency of the proposed calculation method is contrasted and the high exactness with accuracy.

S Fazlul Kareem

Papers

Analysis Of Heart Risk Detection In Machine Learning Using Blockchain