Analytics Practicum Poster Presentations

By: Michael Albuquerque

The stage was set at the CoRE building as the Spring 2017 MBS Analytics students presented posters on their final projects as a part of the Advanced Analytics and Practicum class taught by Dr. Christie Nelson. The class had set up their posters before the event ready to brief professionals from the industry, faculty and fellow students about their work.

Michael Ries, Hungyueh Chang and Yue Sun presenting their project on “Topic Modeling to Predict Movie Success”

Among the wide array of projects on display, Mike Ries and team first took us through their project which included the application of topic modeling to understand and predict the success of movies. Part of the project used an LDA Approach, with parameters such as famous actor count, runtime, international market, famous points and various cultural elements, building a model around existing movie reviews to find hidden patterns to determine what people are talking about. They then used a separate model to predict the commercial success of a movie yet to be released and whether it will win an academy award. The intent behind the model is to aid potential investors to make a decision at the time a movie is being pitched.

Presentations by Jinglei Zhang on Crop Farm Prediction Models for soybeans (left) and Anthony Santos on automated key phrase extraction (right)

Jinglei Zhang spoke about his project to determine metrics for corn and soybean production. He incorporates weather and temperature data into this price prediction model. Predicting these metrics are of great value as they impact producer price index and the consumer price index.

Anthony Santos detailed his approach on a model for automated keyphrase extraction. First eliminating the bulk of words in a publication by filtering out non-nouns and stop words, he went on to significantly reduce the number of candidate words based on other parameters. By adapting his model which he continues to work on, he believes that he will improve the success of this approach involving text modeling.

Zachary Wilkins presents his project on the relation between the case filings of the American Arbitration Association and economic indicators

Zachary Wilkins worked with the American Arbitration Association over the semester on their case filings. Using quarterly data provided, his project was to determine the relation between the case load and economic indicators using a linear regression approach.

A presentation on the Prediction Model for the Australian Real Estate Market by YanJung Zhang

Yanjun Zhang explored the Real Estate Market in Australia. Her project predicted median sales price and growth rate from attributes including population density, internal migration and the value of houses.


Poster Session Networking

Undeniably informative and thought provoking, these projects were a good example of the great work done by our students. In the next article, some MBS Students take us through their work and experiences during the course of their internships. After the short presentations, students were able to network with attendees of the poster session.

See Also: Analytics Students Present Project Results to Employers and Others