This document is about a smart solution for household power management which mainly focused on optimizing the usage of power with cost optimization. In this smart power management system a smart plug and a central device are introduced. Smart plug is for the parameter measurement and central device acts as the main controlling unit. These two devices communicate through wireless technology. These measured parameters are stored in a local database and a remote database. Main tasks like device-identification, pattern recognition and cost prediction are done through analysis of data. The prediction of domestic power consumption and usage optimization are done via data mining and clustering using machine learning techniques. The analysis is done using neural networks, support vector machines, k-means, mean shift and Silhouette classifications. The basic idea is to select a classifier with better performance in real time to detect devices and record the power consumption data set. The device prediction and pattern identification algorithms adapt itself based on the data received via the smart plug. The prediction algorithms have the ability to give expected power consumption of next period and optimization algorithms provide optimized methods to use devices efficiently and effectively.