MScAC (Sep 2016 - present)
Mood-Based Image to Music Synthesis
CSC2541: Differentiable Inference and Generative Models
The goal of this project is to generate music that evokes the same mood as an image. A pre-trained
convolutional neural network extract the features of an image, and is then used to condition the first
timestep of a recurrent neural network. The RNN is trained to produce music.
Analytic Visualization of Book Recommendations
CSC2524: Information Visualization
This project tries to tackle the problem where book recommendations from black-box systems usually
appear in lists that don't facilitate item comparisons. Books are high-dimensional entities that can
be compared on a number of different metrics like popularity and genres. This visualization provides 3
different views that allow users to explore the recommendation space through multiple perspectives.
Proof of Concept
Research Internship (May 2013 - Aug 2013)
SerpenTI: Flexible Analytics of Users, Boards and Domains for Pinterest
The aim of this research project was to provide an analytics platform for Pinterest. Given a query,
we want to provide a list of Pinterest users and boards that are good sources for that query. My
contribution to the research project was the front-end visualization of the analytics.
Honours BSc (Sep 2010 - May 2014)
Specialist in Computer Science
Minor in English
Classification of Tweets
CSC401: Introduction to Natural Language Computing
The goal of this course assignment is to utilize NLP techniques in processing tweets so they can be
classified into categories. Developed in Python, features such as number of adverbs, average word length,
and frequency of punctuation were calculated. Then the WEKA machine-learning module was used to classify
the tweets according to user and/or genre (ex. pop, news, science, etc.)
Classification of Handwritten Digits
CSC411: Introduction to Machine Learning and Data Mining
I implemented three classifiers (k-nearest neighbours, logistic regression and naive Bayes) to
compare their performance in classifying images of handwritten 2s and 3s.
For more information about my Academic background and relevant projects, please download a copy of my