Our next Maritime DevCon 2018 speaker for you to meet is bringing us a session thats all about data. Meet Danielle Leighton. Danielle is a Data Science Strategist at T4G Ltd, where she concentrates on the creation and implementation of industrial artificial intelligence programs with demonstrable impact. She helps clients transform raw sensor data into new insights and data products aligned with their business goals. Undeterred by the messy world of raw sensor data, Danielle is passionate about calibrating sensor data and keeping it in context. Working with architects, data engineers, developers, and visual designers, she adapts approaches and machine learning methods as needed over the life of a project. Her professional background includes energy, retail, healthcare, and government, and she has worked with quantified selfers, biohackers, hacklabs, and makerspaces. She is notoriously unreadable to GSR wearables. In her previous life, Danielle worked with the world’s most sophisticated wearable to date, the hearing aid. Currently, she focuses most of her time on AI programs in the energy sector.
So Danielle, if you were a programming language or technology, which one would you be and why?
My twitter handle, @filatorium, refers to an old French handcraft that forms the lattice-like foundation (network) for fancy lace designs. All of my work hinges on sensors communicating with some collecting program… so I’d like to think if I were a technology, I’d be a communication protocol designed to be that foundation – so cooler stuff can happen.
What do you like best about Maritime DevCon?
I haven’t had the privilege of attending Maritime DevCon (I had a poorly-timed baby last year), but I love that we have a local celebration of our tech talent– that spans the breadth of the technology world. I really believe every technology and language that exists has someone local who works with it, and Maritime DevCon lets us meet and support each other. Other parts of the world have seriously specialized networks, but I think our strength in the maritimes is our flexibility and diversity of talent. Maritime DevCon lets us celebrate that.
Why do you feel your session, Assess & Address: Dealing with Biased Data, will bring value to Maritime DevCon attendees?
This session covers a very underdeveloped area of AI/machine learning/data science. The recent focus has been on exposing the problem of biased source data in our field, but hasn’t begun to develop best practices for dealing with it. Many talks on this subject leave attendees wondering if AI is fundamentally damaging for society (and should be abandoned). Opening this issue up for discussion is critical and does not require a background in AI specifically, but is definitely technical in nature. Maritime DevCon attendees have the right background, and the right desire to be at the forefront of a new wave of best practices to make AI work for us. This issue really requires a multitude of skills and perspectives to be solved- and working as a group across technical and background ‘lines’ is what maritimers do so well!
What is the key take away attendees will get from your session?
This session offers methods to prevent biases from contaminating model results – both at a technical level (for practitioners) and also at a higher level (for all technologists).
Thank you in advance for joining us for Maritime DevCon 2018 Danielle!