The document discusses predictive analysis models that can be used for IoT data from wireless sensor networks. It lists several predictive modeling techniques, including multiple linear regression, support vector machine regression, random forest, gradient boosting machine, and extreme gradient boosting machine. These models are compared to understand their usage for predicting energy use of appliances based on IoT data. IoT data analytics can harness large amounts of structured and unstructured streaming data using these traditional tools and techniques for predictions, identifying trends, finding hidden patterns, and decision making.