Full metadata record
|Department:||Department of Computing||en_US|
|Author:||Wong, Kwan Yeung||-|
|Title:||Visualizing time series data with temporal matching based t-distributed stochastic neighbor embedding||en_US|
|Abstract:||Interpreting time series data has always been a hot research topic for various applications, especially when the dimensionality of time series dataset keeps growing almost prohibitively as technology advances. There exist several dimension reduction techniques attempting to address the related problem, but the inherent nature of time series datasets usually involves factors including time and amplitude shifting and scaling, which could impact the trustworthiness of the visualization results. Moreover, the diversity of domain knowledge also poses another difficulty for this task. For example, same set of time series data could have different kinds of correlation when they are posed in different domains. The implication is that a general similarity metric might not be able to capture such correlation as well as a dedicated metric for time series related applications. Furthermore, the recent recurrent neural networks (RNN) can also be viewed a source of time series data of which the visualization has posted certain requirements but contributed to understand how the RNN works. In other words, a proper understanding of the feature embeddings within the RNN model can provide us more insights on the learning tasks. T-distributed Stochastic Neighbor Embedding (t-SNE) is considered as a highly effective machine learning algorithm for visualization and it is in fact a nonlinear dimensionality reduction technique tailor made to embed high-dimensional data in a low-dimensional space of two to three dimensions only for proper visualization. In view of the key problem of adopting t-SNE to visualize time series data, the project proposes to introduce two temporal matching metrics, namely, dynamic time warping (DTW) and angular metric for shape similarity (AMSS), for t-SNE to enhance its visualization ability for time series data. They provide a more robust similarity metric for time series data so that the embedding process in t-SNE can be made more effective, which are demonstrated by different data visualization experiments. In addition, we also conducted experiments to visualize RNN's activations and the proposed enhancements again show their effectiveness.||en_US|
|Pages:||x, 60 pages : color illustrations||en_US|
|Subject:||Hong Kong Polytechnic University -- Dissertations||en_US|
Files in This Item:
|991022268443203411.pdf||For All Users (off-campus access for PolyU Staff & Students only)||2.38 MB||Adobe PDF||View/Open|
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item: