Yunzhi Zhang

A core problem in developmental research is understanding the mental process of infants and young children underlying the formation of their behaviors. Established experimental paradigms, such as preferential looking, have been essential in decoding these unobservable mental processes from experimental measurements. These paradigms have provided invaluable insights over the years, emphasizing a few low-dimensional metrics for analysis, but the intricacies of infant subjects’ behaviors exist beyond these heuristic metrics. In the past, with classical statistical analysis methods, higher dimensional data are challenging to consume. My research develops machine learning frameworks that leverage artificial neural networks for effective and scalable analysis to harness these data. With the synthesis of methodologies from cognitive science and computer science, I aim to develop a further understanding of human cognitive development.