In the current world data is more important to analyze the surroundings. All the new inventions or future predictions are made using IoT. These data to make our life easy. These days the maximum of data collection is done through IoT, because it is easy and the data can be stored in an organized way. Many countries are working on how to protect this data. The major issue in the world is data security or protection and the topic of research that is going on IoT devices is that these devices should be made that intelligent, so that they can detect useful information and store it on a cloud and storage of inadequate information can be avoided. So, how can we make our IoT devices intelligent. Here machine learning comes into the picture. Machine learning (ML) helps that device to analyze things from the given instructions and store the information on the cloud. ML helps developers to classify the data and clustering/finding pattern in it. With this help, IoT devices can scan the product and analysis of the product using past data or in-build functions. Using Machine learning IoT devices will not only increase the performance of the devices but it will also help in the security of the devices and the data. Whenever a product is related to social welfare or any security-based system, the security of that particular device is given priority because the data that the device is transmitting or receiving may be important or highly confidential. So, if the device is secured, then the leakage of data can be avoided. Apart from this machine learning can be used to detect malicious attacks, analyzing mobile endpoints, a combination of Ml & AI to learn human behavior and to automate wearisome security tasks, and many more. There are many other applications of machine learning in IoT like advancing smart city projects, smart transportations, managing Big Data, thread detection, any kind of accident tracker, and many more smart devices that will help to optimize the remote areas of a country.