IoT and Blockchain for Efficiency in Smart Home Automation with BiLSTM-RNN Model

One of the numerous things made easier by the rapid advancement of technology is home automation, which has been growing in popularity over the past few years. Mechanization and automation have permeated nearly every facet of human existence. Implementing a network that can link different home automations through sensors, actuators, and other data sources is the main objective of this approach. Sensors, preprocessing, feature engineering, and training the model are the three stages that make up the suggested method. Sensors are useful because they can gather information about the environment and then process it to produce results that are correct relative to that environment. The raw data on energy usage is normalized and outliers are removed as part of the preparation steps in the suggested methodology. Label encoding methods were used in feature engineering to convert the input into a feature vector. For training the model, the RNN-BiLSTM-CNN was employed. Compared to alternatives such as RNN and LSTM, the suggested model obtains an average accuracy rate of 95.06%.