This seminar will be held in the LIDA Boardroom 11.87, Worsley Building, at 12.30 on Thursday 19th September.
Seminars are free and open for all to attend. No prior booking is required.
Each presentation will be followed by a short Q&A session.
Presentation 1: Translation Workflow Automation by determining text difficulty
By: Gonzalo Cruz Garcia
What I will be talking about: With increasing globalisation and the push towards automation, it becomes essential to improve on current translation methods. The United Nations (UN) has six official languages and is required to translate an ever-increasing number of documents. In my LIDA internship project, I explore the UN corpus in order to understand the factors contributing to difficulties in both human and machine translation. The time taken for human translation of documents is predicted as well as classification of sentences based on their machine translation difficulty. This is achieved using a combination of Biber dimensions, translation edit rate and sentence vectorisation through Facebook’s XLM neural networks. In September I will be travelling to the UN and World Trade Organisation in Geneva in order to discuss the findings of this project and its applications for improving text translation.
Presentation 2: Classification of primary brain tumours based on their predicted response to standard therapy
By: Stelios Theophanous
What I will be talking about: Glioblastoma (GBM) is the most common, most deadly adult brain cancer. Treatment is standardised and aggressive but almost 100% of tumours recur. Previous research has found that primary GBM tumours that undergo standard treatment exhibit a universal dysregulation in expression of genes associated with a specific chromatin remodelling protein. However, the direction of this dysregulation is patient-specific. Our hypothesis is that this epigenetic switch facilitates the inevitable treatment resistance of GBM tumours. The aim of this project is to develop a machine learning classifier that predicts the direction of the epigenetic switch based on the gene expression profile of the primary tumour. Stratifying patients according to the way their tumour will switch may ultimately determine a treatment course which will more effectively kill the tumour.
Presentation 3: Data Assimilation for Agent Based Models using Probabilistic Programming
By: Benjamin Wilson
What I will be talking about: This project explores the application of modern probabilistic programming languages on agent tracking and prediction.
Civil emergencies such as flooding, terrorist attacks, fire, etc., can have devastating impacts on people, infrastructure, and economies. Knowing how to best respond to an emergency can be extremely difficult because building a clear picture of the emerging situation is challenging with the limited data and modelling capabilities that are available. Agent-based modelling (ABM) is a field that excels in its ability to simulate human systems and has therefore become a popular tool for simulating disasters and for modelling strategies that are aimed at mitigating developing problems. However, the field suffers from a serious drawback: models are not able to incorporate up-to-date data (e.g. social media, mobile telephone use, public transport records, etc.). Instead they are initialised with historical data and therefore their forecasts diverge rapidly from reality.
To address this major shortcoming, this research project will develop dynamic data assimilation methods for use in ABMs with the use of Probabilistic Programming Languages. There are serious technical and methodological barriers that must be overcome, but this research has the potential to produce a step change in the ability of models to create accurate short-term forecasts of social systems.