IoT devices are various types of hardware such as appliances, sensors, machines, or actuators that are programmed for specific applications such as data transmission over a network or the Internet. They are embedded in other industrial devices, biological sensors, mobile devices, and medical devices. There are approximately 7.62 billion people in our world, with a growing graph of IoT devices. So, the amount of data released from these IoT devices also increases, and there may be a chance of leakage of data from these devices. These IoT devices used in our home is exposed to different types of attacks like eavesdropping, brute force attack, leakage of information, cyber attacks etc. In detecting these types of attacks the machine learning algorithms play an important role and which improves the security of these IoT devices and help in making these IoT devices more secure; that is the primary goal of this project. Machine learning models can help in finding the spam in the data. Smart home datasets are available, which includes different weather conditions is used for finding spam. The main models like Long Short-Term Memory (LSTM), Time Series and Recurrent Neural Network are used in finding the spam in the dataset. The proposed model is able to find out the spam in the data of the IoT devices.