Matching Items (5)
Filtering by
- All Subjects: Probability
- Creators: Wilson, Jeffrey
- Creators: Goegan, Brian
- Member of: Barrett, The Honors College Thesis/Creative Project Collection
- Resource Type: Text
Description
Predictive analytics have been used in a wide variety of settings, including healthcare,
sports, banking, and other disciplines. We use predictive analytics and modeling to
determine the impact of certain factors that increase the probability of a successful
fourth down conversion in the Power 5 conferences. The logistic regression models
predict the likelihood of going for fourth down with a 64% or more probability based on
2015-17 data obtained from ESPN’s college football API. Offense type though important
but non-measurable was incorporated as a random effect. We found that distance to go,
play type, field position, and week of the season were key leading covariates in
predictability. On average, our model performed as much as 14% better than coaches
in 2018.
sports, banking, and other disciplines. We use predictive analytics and modeling to
determine the impact of certain factors that increase the probability of a successful
fourth down conversion in the Power 5 conferences. The logistic regression models
predict the likelihood of going for fourth down with a 64% or more probability based on
2015-17 data obtained from ESPN’s college football API. Offense type though important
but non-measurable was incorporated as a random effect. We found that distance to go,
play type, field position, and week of the season were key leading covariates in
predictability. On average, our model performed as much as 14% better than coaches
in 2018.
ContributorsBlinkoff, Joshua Ian (Co-author) / Voeller, Michael (Co-author) / Wilson, Jeffrey (Thesis director) / Graham, Scottie (Committee member) / Dean, W.P. Carey School of Business (Contributor) / Department of Information Systems (Contributor) / Department of Management and Entrepreneurship (Contributor) / Barrett, The Honors College (Contributor)
Created2019-05
Description
Predictive analytics have been used in a wide variety of settings, including healthcare, sports, banking, and other disciplines. We use predictive analytics and modeling to determine the impact of certain factors that increase the probability of a successful fourth down conversion in the Power 5 conferences. The logistic regression models predict the likelihood of going for fourth down with a 64% or more probability based on 2015-17 data obtained from ESPN’s college football API. Offense type though important but non-measurable was incorporated as a random effect. We found that distance to go, play type, field position, and week of the season were key leading covariates in predictability. On average, our model performed as much as 14% better than coaches in 2018.
ContributorsVoeller, Michael Jeffrey (Co-author) / Blinkoff, Josh (Co-author) / Wilson, Jeffrey (Thesis director) / Graham, Scottie (Committee member) / Department of Information Systems (Contributor) / Department of Finance (Contributor) / Barrett, The Honors College (Contributor)
Created2019-05
Description
We attempted to apply a novel approach to stock market predictions. The Logistic Regression machine learning algorithm (Joseph Berkson) was applied to analyze news article headlines as represented by a bag-of-words (tri-gram and single-gram) representation in an attempt to predict the trends of stock prices based on the Dow Jones Industrial Average. The results showed that a tri-gram bag led to a 49% trend accuracy, a 1% increase when compared to the single-gram representation’s accuracy of 48%.
ContributorsBarolli, Adeiron (Author) / Jimenez Arista, Laura (Thesis director) / Wilson, Jeffrey (Committee member) / School of Life Sciences (Contributor) / Barrett, The Honors College (Contributor)
Created2021-05
Description
This paper proposes that voter decision making is determined by more than just the policy positions adopted by the candidates in the election as proposed by Antony Downs (1957). Using a vector valued voting model proposed by William Foster (2014), voter behavior can be described by a mathematical model. Voters assign scores to candidates based on both policy and non-policy considerations, then voters then decide which candidate they support based on which has a higher candidate score. The traditional assumption that most of the population will vote is replaced by a function describing the probability of voting based on candidate scores assigned by individual voters. If the voter's likelihood of voting is not certain, but rather modelled by a sigmoid curve, it has radical implications on party decisions and actions taken during an election cycle. The model also includes a significant interaction term between the candidate scores and the differential between the scores which enhances the Downsian model. The thesis is proposed in a similar manner to Downs' original presentation, including several allegorical and hypothetical examples of the model in action. The results of the model reveal that single issue voters can have a significant impact on election outcomes, and that the weight of non-policy considerations is high enough that political parties would spend large sums of money on campaigning. Future research will include creating an experiment to verify the interaction terms, as well as adjusting the model for individual costs so that more empirical analysis may be completed.
ContributorsCoulter, Jarod Maxwell (Author) / Foster, William (Thesis director) / Goegan, Brian (Committee member) / School of Mathematical and Statistical Sciences (Contributor) / Department of Economics (Contributor) / Dean, W.P. Carey School of Business (Contributor) / Barrett, The Honors College (Contributor)
Created2019-05
Description
This paper analyzes responses to a survey using a modified fourfold pattern of preference to determine if implicit information, once made explicit, is practically significant in nudging irrational decision makers towards more rational decisions. Respondents chose between two scenarios and an option for indifference for each of the four questions from the fourfold pattern with expected value being implicit information. Then respondents were asked familiarity with expected value and given the same four questions again but with the expected value for each scenario then explicitly given. Respondents were asked to give feedback if their answers had changed and if the addition of the explicit information was the reason for that change. Results found the addition of the explicit information in the form of expected value to be practically significant with ~90% of respondents who changed their answers giving that for the reason. In the implicit section of the survey, three out of four of the questions had a response majority of lower expected value answers given compared to the alternative. In the explicit section of the survey, all four questions achieved a response majority of higher expected value answers given compared to the alternative. In moving from the implicit to the explicit section, for each question, the scenario with lower expected value experienced a decrease in percentage of responses, and the scenario with higher expected value and indifference between the scenarios both experienced an increase in percentage of responses.
ContributorsJohnson, Matthew (Author) / Goegan, Brian (Thesis director) / Foster, William (Committee member) / School of Sustainability (Contributor) / Economics Program in CLAS (Contributor) / Dean, W.P. Carey School of Business (Contributor) / Barrett, The Honors College (Contributor)
Created2019-05