Identified characteristics of social media users, namely gender and age, by training deep learning (including vanilla RNN and LSTM) and classical machine learning models (including SVM and logistic regression) on the textual posts they have shared on Twitter.
- Achieved 79.6% validation accuracy with a feedforward neural network consisting of word embedding, flattening, and two densely-connected layers.
- Achieved 76.9% accuracy with a bidirectional LSTM architecture.
A thesis submitted in conformity with the requirements for the degree of:
Master of Science
School of Electrical Engineering and Computer Science
Faculty of Engineering
University of Ottawa
Research funded by a grant from Natural Sciences and Engineering Research Council of Canada (NSERC).