Tunable Floating-Point for Embedded Machine Learning Algorithms Implementation | IEEE Conference Publication | IEEE Xplore