ML Retreat 2017

(I wish this had a title)

Tania Allard

2nd June, 2017

About me

  • I completed my PhD in Manchester ( Whoop!!! )
  • I am a member of the RSE (Research Software Engineering) team
  • My interests: research reproducibility, data engineering/science, community building, optimisation, diversity in STEM areas

...I love being a RSE

  • We help researchers to develop robust, reliable, well documented/structured software because not only data is key ... so is your software
  • We are involved in mentoring, community building, and outreach activities
  • Even though we are geeks... we are social geeks
  • We get to get our hands on loads of cool stuff

Code First: girls

A social initiative to promote the integration of women in tech-related careers

RSE conference

Either within the organising committee, as presenters, or just attending.

Data Study Group

The Alan Turing Institute (22-26/05/17)

RA's, PhD's and people from various industries worked together to tackle some industrial 'Data Science / Machine Learning challenges':
  • Predict events location based on geographical data and historical events data (DSTL)
  • Identify and predict cyber attacks from traffic data (DSTL)

Thomson Reuters challenge

The joy of working in teams

  • People from absolutely different backgrounds
  • People with completely random skills
  • Very limited time to complete a 'big-project': from data wrangling to report writing
  • Loads and loads of data to analyse (which needed to be scraped beforehand)

A minimal idea of the data

More team working joy

As a PhD In a diverse team
Ignore version control :( Need version control
Local work? Sharing? Efficient content sharing approach
Self-directed Collaboration is key
Choose preferred tools Choice / adapatation
Thesis structure/ content matters Every bit matters