Note: Please use your full real names where signing up, otherwise we have problems with building security.
Kevin Keraudren (https://github.com/kevin-keraudren) on Medical Imaging in Python: from morphological operations to machine learning
In this talk, I will present how to leverage the Python ecosystem to build complex analysis pipelines for medical image analysis. Examples will be drawn from my time at Imperial College, working on microscopy images (http://www.slideshare.net/kevinkeraudren/segmenting-epithelial-cells-in-highthroughput-rnai-screens-miaab-2011) and fetal MRI (https://www.youtube.com/watch?v=t1pIh_TOVLM), as well as the work we are currently doing at Klarismo (https://klarismo.com/) on quantitative analysis of whole body MRI. This talk will involve scikit-image for image processing, scikit-learn for machine learning, SimpleITK for algorithms specific to medical images, and VTK for 3D visualisation. And because medical images tend to have lots of voxels, it will surely require some Cython code to speed things up.
Nicholas Radcliffe (http://www.tdda.info/) on Test-Driven Data Analysis
Test-driven data analysis (TDDA) seeks to take some of the ideas and practices from test-driven development and apply them, with appropriate modifications, to avoid some of the common pitfalls in data analysis. While some of the ideas carry over straightforwardly, both the way data analysis is typically developed and the nature of data analysis tasks mean that mainstream TDD ideas need to be modified and extended to be fully effective in the context of data analysis.
This talk will compare and contrast mainstream TDD and TDDA, and talk about the challenges and ideas for developing further this important new area. It will also show some open source libraries we are developing to support TDDA. Note: For anyone who attended PyCon UK 2016, this abstract is very similar to that talk, but the content and libraries have moved on considerably since Cardiff.
Nicholas Radcliffe (http://tdda.info) on Rexpy
Most of us love the power of regular expressions, but dislike writing them and struggle reading them. Rexpy is an approach to (and library for) reducing the pain in writing regular expressions.
Miroslav Batchkarov (http://twitter.com/loglinear) on Page Rank and its applications to keyword extraction and document summarization
This talk introduces the PageRank algorithm and several ways of using it for natural language processing. These include keyword extraction, document summarisation, semantic similarity and word sense disambiguation.
Lukasz Bonenberg (https://twitter.com/LKBLab) on How to get best of Android GPS in python
Android 7.x Nougat introduces raw GPS ranges. In this lightning talk, I will demonstrate how to calculate them and what is the benefit of using them instead of current "black box" position output. I am also interested in finding collaborators for rewriting existing Matlab code (https://github.com/google/gps-measurement-tools) in python.
Doors open at 6.30 (get there early as you have to sign-in via AHL's security), talks start at 7pm, beers from 9pm in the bar. We normally have > 200 folk in the room so there's plenty of people to discuss data science questions with!
Please unRSVP if you realise you can't make it. We're limited by building security on number of attendees, so please free up your place for your fellow community members!
Follow @pydatalondon (https://twitter.com/pydatalondon) for updates and early announcements. See you on the 1st!