A comprehensive evaluation of stacked autoencoders for text embedding

Bhatnagar, Sarthak (2023) A comprehensive evaluation of stacked autoencoders for text embedding. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (1MB) | Preview
Preview	PDF (Configuration manual) Download (1MB) | Preview

Abstract

Artificial Intelligence is significantly expanding in the current age of Information Technology, with many people utilizing the potential of AI to make machines intelligent enough to perform repetitive tasks creatively and more simply. With the advancement of technologies, AI research, and development have narrowed to two research areas: Machine Learning and Deep Learning. Machine learning algorithms are trained on data to learn how to perform a task. However, they are stringent for domain knowledge, while deep learning (a subset of machine learning) has an excellent capability to achieve flexibility in computational tasks and accuracy and offers to learn through neural network architecture. In this research paper, we examine the effect of a stacked autoencoder architecture on text embedding by employing Bidirectional Long Short-Term Memory (BiLSTM) and Bidirectional Gated Recurrent Unit (BiGRU) models. The stacked autoencoder architecture can capture complex patterns and parameters within the dataset due to multiple layers and Bidirectional LSTMs and Bidirectional GRUs, known for their capability to capture context from both past and future sequences, are used as components to study autoencoder architectures. By leveraging the strengths of both, we will show that there is indeed a trade-off between model complexity and accuracy using layered architecture. Our results showcased that a three-layered bidirectional GRU autoencoder has the best accuracy. Moreover, the higher number of layers has a negligible impact on the accuracy, while potentially taking more computing resources.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Estrada, Giovani UNSPECIFIED
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science Q Science > QA Mathematics > Algebra > Algorithms > Computer algorithms Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Tamara Malone
Date Deposited:	08 Nov 2024 12:34
Last Modified:	08 Nov 2024 12:34
URI:	https://norma.ncirl.ie/id/eprint/7172

Actions (login required)

View Item