Lung cancer prediction

In the United States, there are approximately 225 cases of lung cancer each year. It is estimated that the cost of healthcare is $ 000 trillion annually. In 12, the US Vice President's office launched the "Cancer Moonshot" initiative to expand lung cancer prevention, early detection and treatment programs.

In 2017, to support this initiative, a "Data Science Bowl" tournament was organized, which combined the medical community with data analysts to create algorithms for lung cancer detection. This year, the prize total was $ 1 million. The team (IBM Kraków data science team) tested a wide range of technologies, including deep learning tools such as Tensorflow / Keras, Caffe, Theano, as well as experimented with machine learning (Scikit-learn, XGBoost and SparkMLlib). As part of this presentation, we would like to share our experience gained during this tournament and explain how Watson Machine Learning service was used to prepare the data, build the model and the learning itself.

Purpose / nature of the offer:

  1. Promotion of IBM Kraków and spreading knowledge of what we do
  2. Promotion of the Watson Machine Learning product
  3. Spreading knowledge about machine learning

Meet the presenter:

Łukasz Ćmielowski

Łukasz Ćmielowski

He joined IBM in 2008. In 2009, he defended his doctorate in the field of DNA microarray data analysis at the Silesian University of Technology. Since then, he has worked as a QA architect in one of the IT infrastructure management products. He focused on the automation of test tasks and the use of learning machines to predict errors in the software. In 2015, he joined a new solutions team from the analytics group as an architect. He is a huge fan of books by Norman Davis and Terrry Pratchett.

We invite you to listen to the entire presentation during this year's edition
Life Science Open Space.
19.10.2016, godz. 9:00-20:00
ICE Congress Center ul. Maria Konopnicka 17 Cracow