In the dissertation we provide the quantitative methods for analyzing and understanding the co-feature and innovation in the creation of Western classical music by using complex systems theories. In order to observe the co-feature landscape of classical musicians, we build the co-feature network of classical musicians from the dataset of classical music recordings. Then we analyze its topology and dynamics and propose a multi-scale analysis which enables us to investigate the co-feature topology in more detail. We also propose a method for measuring the innovativeness and influence of composers by modeling music with codewords to better understand the innovation and development in classical music. From the co-feature of classical musicians we find its distinguishing structural characteristics and the co-existence of universality and diversity, two seemingly paradoxical phenomena, in classical music; while a minority of prominent musicians dominate the classical music industry, musical diversity is preserved by the heterogeneous co-feature structures of groups and individuals. The quantitative analyses of the innovativeness and influence of composers help us to understand the association between the musical aesthetics and innovativeness and the generation and development of music styles from the Baroque to the Romantic era. This study is applicable to the quantitative study of the co-feature, or generally collaboration, and innovation in other cultural systems. Furthermore we believe that it would be of a great help to scientifically understand complex values such as creativity and success manifested through collaboration and innovation in culture.