This chapter summarizes three key concepts to be used in machine learning methods. Therefore, we will first start with random numberRandom number generation as pseudosensor data. This will help us while testing machine learning methods when there is no active sensor around. Then, we will consider feature extractionFeature extraction as a way of summarizing the acquired sensor data as the second concept. This will act as a bridge between Chap. 4 on data acquisitionData acquisition and the following chapters. Then, we will evaluate sensor data normalization as the third concept. Finally, we will consider feature extractionFeature extraction from accelerometerAccelerometer data, audio signals,Audio signal and digital imagesDigital image as end of chapter applications. These will show us how the topics considered in this chapter can be applied to real-life machine learning problems.