Providing small medium business with data science capabilities. 

Data Scientist

I use Python as my main tool as a Data Scientist. As part of my PhD, I tutored Python to several students from all levels of coding.

My PhD revolved around NLP, in particular sentiment analysis. My main focus was to demonstrate that idiomatic phrases, which are often ignored in traditional approaches, have a significant impact on the overall results of sentiment analysis. I also have a keen interest in word sense disambiguation, text and sentiment classification, data annotation, and pre-processing natural language text. I'm also very experienced in using WordNet.

I enjoy teaching NLP to those of all levels. 

I use and apply machine learning for several business and real problems. My main experience is in text classification, but I also have strong experience in the feature engineering, data processing and classification of cybersecurity related data.

I'm currently in the process of self-learning time series forecasting.

I have 5 years of experience in tutoring University students databases using SQL. 

I used Weka thoroughly as part of my PhD to perform machine learning experiments.

BSc, PhD, Data Scientist

Lowri Williams

Here's a list of the academic literature that I have published.

Academic publications

Here's a piece that I wrote about my analysis on the coronavirus data.



🦠 Coronavirus: A Data Scientist's Perspective

Here's a piece that I wrote about the development and analysis of my Spotify Sentiment Analysis app.

🎶 Spotify Sentiment Analysis

Accessing and analysing Six Nations Rugby data, as well as predicting the scores of the remaining postponed matches.

🏉  Predicting the Scores of the Postponed 2020 Rugby Six Nations Matches

An investigation into how to assign topics with meaningful titles.



🗣️ Topic Modelling: Going Beyond Token Outputs

Analysing and classifying real patient cough sounds.


🦠 Coronavirus: Using Machine Learning to Triage COVID-19 Patients

Exploring how adversarial learning can be used to target SMS spam detectors using Jacobian-based Saliency Map attacks

Personal details

Lowri W.

Summary

Technical skills

Work Experience

Personal Projects