In this talk, we first review existing results of quantum machine learning.
Second, we discuss a crucial parameter in machine learning - the sample complexity, which determines the number of queries to a membership made by the learning algorithm such that the hypothesis function is Probably Approximately Correct. Third, we provide a framework to analyze learning matrices in the Schatten class by taking advantage of matrix concentration inequalities. As a result, we establish the fat-shattering dimension of learning bounded operators and trace class operators. By characterizing the tasks of learning quantum states and two-outcome quantum measurements into learning matrices in the Schatten-1 and 1 classes, our proposed approach directly solves the sample complexity problems of learning quantum states and quantum measurements.