There is a difference between data and information. In most cases, the humongous amount of data contains only concise information. Broadly speaking, compressed sensing deals with duality—abundance of data and its relatively sparse information content. Truly speaking, compressed sensing is concerned with an important sub-class of such problems—where the sparse information content has a linear relationship with data. There are many such problems arising in real life. This chapter discusses a few of them. It broadens the definition of “compressed sensing” and includes low-rank matrix recovery as well. The chapter illustrates some of the problems in machine learning and signal processing involving functional magnetic resonance imaging.