Abstract
Machine learning (ML) techniques will benefit immensely from the avalanche of data readily available from various (IoT) applications considered as the major contributor of new data for future intelligent network. Based on this new concept, network systems will further magnify their capacity to exploit variety of experimental data across a plethora of network devices, study the data information, obtain knowledge and make informed decisions based on the dataset at their disposal. Smart IoT data analysis are performed utilizing supervised learning, unsupervised learning and reinforced learning. This study is limited to supervised and unsupervised ML techniques. In other to achieve the set objectives, reviews and discussions of substantial issues related to supervised or unsupervised machine learning techniques were executed, highlighting the advantages and limitations of each algorithm as well presenting the recent research trends and recommendations for future study.