Predictive Modeling for the Game of Twilight Imperium to Develop Winning Strategy Ian Seibel Project Mentor(s): Yvonne Chueh, PhD In this presentation, we look at the board game Twilight Imperium, created by Fantasy Flight Games, and attempt to discover how typically players have been able to win within the asymmetric nature of the game. Using data taken from a multitude of online games, we look at the different factors that might have caused a player to win over another. We use a variety of predictive models, such as related regression, random forests, and neural networks to determine what factors have the biggest impact on the final scores of the game. We also observe general and faction specific strategies that optimize the number of points a player can earn. Presentation Type: Oral Presentation (May 20, 9:30am–5:00pm) Keywords: Optimization, Predictive Modeling, Probability Analysis, Game Theory SOURCE Form ID: 162 Car Insurance Premium Affected by Tariff and Macroeconomic Factors Hang Su, Cade Pratz Project Mentor(s): Yvonne Chueh, PhD This research project develops a predicative model for auto insurance premiums using macroeconomic and specific vehicle factors over the past decade. By using serval key variables to predict future insurance costs including Consumer Price Index (CPI), Tariffs, Actual Cash Value for vehicles (ACV), and Commercial and Private Vehicle Insurance costs; by using the up-to-date data from Federal Reserve Economy Data (FRED) to introduce a new model Tariff Inflation Pressure Index (TIP), to present how did the vehicle insurance costs response to the model. This Washington State of Transportation (WSDOT) and evaluated across the broader model is trained on real world data from dataset to assess its ability to generalize across different vehicle categories. To identify the most impactful drivers of insurance costs and to determine how well macroeconomic indicators can explain variations in premiums to improve the accuracy of the data, to minimize the impact of driving behavior, then analyze the data. This approach provides a simplified by practical framework for understanding and predicating insurance pricing behavior across different segments of the automotive market. Presentation Type: Oral Presentation (May 20, 9:00am–5:00pm) Keywords: Vehicle Insurance, Tariff, Inflation SOURCE Form ID: 241
138
Made with FlippingBook interactive PDF creator