## Cillian Hourican, PhD Candidate

## About me

Hi! I`m a PhD student in Trinity College Dublin, in the School of Computer Science and Statistics.

In May 2018, I graduated from TCD with a first-class honours degree in Mathematics. In September 2018, I started my PhD in Computer Science under the supervision of Prof. Douglas Leith.

houricc@tcd.ie Lloyd Institute

Trinity College Dublin

Ireland

## Research Interests

Broadly speaking, my research is in the area of mathematical optimisation, machine learning and statistical learning. I am also fascinated by the area of Deep learning. My current focus is on convex optimisation and online learning.

I am also taking classes on Deep Learning, with a focus on image classification, feedforward neural networks, convolutional neural networks, network architecture, test processing, RNNS, GANS and Autoencoders, making use of the Keras package in Python.

I am also enrolled in classes on Optimisation algorithms for Data Analysis, Applied Statistical Modelling, Research Integrity and Research Methods, with the aim of further enhancing my research abilities.

## Teaching

I am a Teaching Assistant and/or Demonstrator for the following courses:

- ST1251 - Introduction to Statistics 1
- ST2005 - Applied Probability 2
- MAU11001 - Mathematics, Statistics and Computation
- MAU11002- Mathematics, Statistics and Computation

## Education

### PhD Candidate

Trinity College Dublin

### B.A Mathematics

Trinity College Dublin

### Leaving Certificate

Moyne Commnity School

## Work Expericnce

### Teaching Assistant / Demonstrator - Probability and Statistics

Trinity College Dublin

Give lab tutorials to first year science students, where I show students how to use the R programming language to use the statistical methods they cover in lectures. Also give weekly tutorials to Mathematics students taking probability modules, where I explain solutions to students, encourage class discussions and answer students’ questions.

### Summer Research Student

Trinity College Dublin

Investigated Changepoint Detection Algorithms using both Frequentist and Bayesian Approaches. The analysis was restricted to when only one changepoint in the timeseries was present. Maximum likelihood estimates were compared to estimates found using a Gibbs sampler in the Bayesian setting.

Replicated results from a research paper that focused on a changing linear regression model, where the parameters of the model changed. This made extensive use of the open source programming software R.

### Summer Research Student

LM Ericsson Ireland

Read and summarised research papers related to Self-Organising Networks, researched topics in graph theory for machine learning, took part in team meetings to add our ideas to the main project, learned about several types of graphs and how they are used in statistical and modelling problems.