Measuring the attributes of software is a pivotal umbrella activity throughout the development process of software. The accuracy of measurements controls the successful completion of the project within the pre-decided timeframe and money constraints. With the changing needs of the current software industry, it is desirable to have advanced techniques apt for accurate measurements of software in alignment to the grown complicated software systems. Machine Learning (ML) techniques have been proven to be effective for accurate measurements in past years of research. Further, it needs enhancements to serve with more effective measurements of metrics associated with different tasks and activities of software development. This chapter aims to elaborate on the existing methodologies for software measurements using machine learning techniques with the intent to gain insight into the techniques being deployed in literature. Another major contribution is to elaborate on the tasks from the software measurement domain that can be modeled as machine learning tasks. It also highlights the most suitable machine learning model from the existing literature for software. This chapter will serve as an enlightenment for the academicians and researchers who are willing to work in the domain of applications of machine learning in software development.