Learning parameterized policies for Markov decision processes through demonstrations | IEEE Conference Publication | IEEE Xplore