We started the week with an introduction to machine learning. Our data analysis engineer, Matilda Landgren, introduced the team to the foundation of machine learning. How can our system, the eHealth-platform LifePod, in the future learn about certain features to predict certain patient outcomes, was one question discussed during the session. Matilda also gave us a walk through the most common methods in machine learning, like kNN-classifier, logistic regression, random forests and neural networks, which was highly appreciated by the entire team.

Used in a smart way, machine learning has incredible possibilities to detect and predict non-adherence patterns and to help the health care system to give high quality care to patients. The day we have more data and apply well working machine learning methods, we might be able to predict serious conditions with high accuracy which we wouldn’t have seen with traditional methods. It is still a long way to go but it is always good trying to think ahead.

Text and photo: Cross Solutions,
cathrin.jung@cross-solutions.com