Document Type
Dissertation
Date of Award
12-31-2024
Degree Name
Doctor of Philosophy in Computing Sciences - (Ph.D.)
Department
Computer Science
First Advisor
Ioannis Koutis
Second Advisor
Amy K. Hoover
Third Advisor
Guiling Wang
Fourth Advisor
Shantanu Sharma
Fifth Advisor
Lijing Wang
Sixth Advisor
Mark J. Nelson
Abstract
The dissertation draws inspiration from the topic of peer learning in the social sciences and the study of information dissemination and knowledge diffusion in network science. In particular, it introduces and studies a setting involving a population or network of artificial learners, with the objective of optimizing aggregate performance measures under constraints on training resources. In this context, natural knowledge diffusion processes in networks of interacting artificial learners are studied. The term "natural" refers to processes that emulate human peer learning, where the internal state and learning processes of students remain largely opaque, and the main degree of freedom lies in the formation of peer learning groups by a coordinator who can evaluate learners before assigning them to peer groups. Among else, empirical findings demonstrate that natural knowledge diffusion (a) occurs despite the presence of misinformation, even in challenging network topologies, (b) effectively utilizes training resources, and (c) supports the design of modular neural models capable of generalization without overfitting noisy labels. Addressing the original question of peer learning, the dissertation evaluates the outcomes of various policies for forming peer groups and identifies slight variations in outcomes as predicted by simple analytical models. The contributions of this work reveal phenomena that are of central importance to Machine Learning and Artificial Intelligence while highlighting the potential of using AI and abundant computational resources as a new empirical lens for examining significant social phenomena.
Recommended Citation
Beikihassan, Ehsan, "Knowledge diffusion in networks of artificial learners" (2024). Dissertations. 1801.
https://digitalcommons.njit.edu/dissertations/1801
