This thesis investigates the viability of an algorithm designed to identify hidden nodes within a polytree structure, subject to specific degree conditions. We evaluate the algorithm’s efficacy through the generation of synthetic data using Gaussian Structural Equation Models (SEMs), initially implementing it based on the example provided in the foundational paper. Subsequent simulations, encompassing variations in sample size, tree size, and significance levels for t-tests of correlation matrices, reveal notable improvements in algorithm performance with increasing sample size. However, notwithstanding these advancements, certain limitations impede the algorithm’s perfection. While the preliminary findings hold promise, further investigation is warranted to enhance the algorithm’s performance.     «

This thesis investigates the viability of an algorithm designed to identify hidden nodes within a polytree structure, subject to specific degree conditions. We evaluate the algorithm’s efficacy through the generation of synthetic data using Gaussian Structural Equation Models (SEMs), initially implementing it based on the example provided in the foundational paper. Subsequent simulations, encompassing variations in sample size, tree size, and significance levels for t-tests of correlation matric...     »