Friday, June 19 2020
10:00am - 12:00am
Masters Presentation
MS Presentation Siyuan Lin

Title: Application of Machine Learning Models to Microeconomic Analysis

Committee: Laura Argys, Erin Austin, Yaning Liu (Advisor and Chair)

Abstract:
This paper briefly introduces the history, evolution and application of machine learning techniques along with the concepts of supervised learning, unsupervised learning, semi-supervised learning and reinforcement learning algorithms. In detail, it explains the mathematical concepts behind fully connected neural networks algorithms and K-means clustering algorithms. Empirically, supervised learning results of the traditional regression analyses and fully connected neural networks are compared and examines K-means clustering analysis is conducted. The data were collected from the American Community Survey between 2005 and 2018 and have been restricted to male Asian immigrants to the US from the five largest sending countries. By applying the same analytical techniques to different combinations of attributes, evidence has been obtained suggesting that under correct specification, additional attributes do not increase fitting accuracy for predictive models. Adding more characteristics to the model only increases fit when the regression specification is used. Using the Elbow method to split data into clusters, possible existence of hidden correlations among observations other than noticeable characteristics has been found.
Speaker:Siyuan Lin
Affiliation:
Location:Zoom meeting


Download as iCalendar

Done