Ge, Xin

Increasingly, artificial neural networks are explored to learn relationships among temporal sequence data for purposes of classification, prediction, and anomaly detection with the hope of exceeding the performance of more traditional machine learning algorithms. While the underlying Long Short-Term Memory or Gated Recurrent Unit networks are still the preferred choices by many researchers, such recurrent networks are sub-optimal to learn relationships within and across longer sequences. Transformer neural networks, originally designed to improve the performance of natural language processing tasks, pose an interesting alternative as their attention mechanisms are more capable of capturing context and meaning within longer sequences. Such features present opportunities to apply transformer networks also to temporal sequence data of financial asset prices. This thesis introduces an extension of the original transformer neural network which is capable of multivariate time series representation learning in a supervised learning context and attempts to train temporal sequences of financial asset prices. The prediction accuracy of the transformer extension exceeds two of the most popular recurrent neural networks used for temporal sequence data prediction. The experiments are conducted in the context of a trading algorithm that showcases the practical potential and its implications. As the model is not input data specific, opportunities to transfer enhancements to other domains exist.

Consumer evaluations of brand extension are becoming increasingly important to the consumer market (Kaur & Pandit, 2015); however, little attention has been given to crosscategory specific research in this field. This research examines whether there are correlations between an iconic product (a product category already occupied by the brand) and its crosscategory extension product and how the user experience on an extension product affects its iconic product. The findings reveal that consumers have a positive attitude toward an extended product when they perceive credibility, quality, and innovativeness from its iconic product. Consumer perceived image-fit and advertisement-match are positively correlated with consumer attitude. The results support that the post-evaluation on an extended product affects its iconic product; however, user experience with an extended product does not correlate with consumers' evaluations of an iconic product on their evaluations of the extended product because of the survey limitation.

Using a Construal Level Theory (CLT) foundation, the authors conduct four studies which find consumers are more likely to pay attention to short-term (long-term) benefits if an event is taking place in the near (distant) future. Additionally, when people are deciding for themselves (acquaintances), they’re more likely to pay attention to short-term (long-term) benefits and proximal (distant) spatial locations. This research provides theoretical and managerial implications, as businesses can tailor marketing campaigns to emphasize short-term/long-term attribute dimensions to prime consumers to choose a certain alternative depending on how psychologically distant they are from an event/object. The research methods used were questionnaires where participants chose between two alternatives. The current research aims to uphold philosophy from previous literature that states: a primary aim of consumer research is to understand aspects that are influencing different trade-offs of a choice set in the preference construction process (Bettman, Luce, & Payne, 1998).