In the last decade, Human Activity Recognition (HAR) has become a very important research area, especially due to the spread of electronic devices such as smartphones, smartwatches and video cameras present in our daily lives. In addition, the advance of Deep Learning (DL) has led researchers to use HAR in various domains including health and well-being applications. HAR is one of the most promising assistive technology tools to support elderly's daily life. However, this class of algorithms requires large amounts of data. Furthermore, not all the HAR application fields generate a significant amount of data and not all of them provide the computational power that DL models in HAR require. This survey focuses on critical applications of Machine Learning (ML) in the fields of HAR, largely oriented to Daily Life Activities by presenting an overview of the publications on HAR based on ML and inertial, physiological and environmental sensors.