No abstract

Nowadays, data generation is huge in nature and there is a need for analysis, visualization and prediction. Data scientists find many difficulties in processing the data at once due to its massive nature, unstructured or raw. Thus, feature extraction plays a vital role in many applications of machine learning algorithms. The process of decreasing the dimensions of the feature space by considering the prime features is defined as the dimensionality reduction. It is understood that with the dimensionality reduction techniques, redundancy could be removed and the computation time is decreased. This work gives a detailed comparison of the existing dimension reduction techniques and in addition, the importance of Principal Component Analysis is also investigated by implementing on the animal classification. In the present work, as the first phase the important features are extracted and then the logistic regression (LR) is implemented to classify the animals.

No abstract

Big data and IoT has its impact on various areas like science, health, engineering, medicine, finance, business, and mainly, the society. Due to the growth in security intelligence, there is a requirement for new techniques which need big data and big data analytics. IoT security does not alone deal with the security of the device, but it also has to care about the web interfaces, cloud services, and other devices that interact with it. There are many techniques used for addressing challenges like privacy of individuals, inference, and aggregation, which makes it possible to re-identify individuals' even though they are removed from a dataset. It is understood that a few security vulnerabilities could lead to insecure web interface. This chapter discusses the challenges in security and how big data can be used for it. It also analyzes the various attacks and threat modeling in detail. Two case studies in two different areas are also discussed.

No abstract