This post concerns building reliable artificial intelligence (AI) systems that we can trust. We will look at reliable AI with respect to a specific application: the Child Growth Monitor (CGM), an AI-driven app to detect malnutrition among children.
Building a reliable AI system cannot be achieved by a single method alone. Only multiple approaches combined allow to set up a reliable system. We will discuss some approaches that directly look at data, as well as different approaches that analyze model predictions. We will also study the monitoring of this IT system and organizational methods.
The author works as an AI…
If you are a researcher in computer science, you know the following situation: You find a paper on arXiv, you read it, and you like it a lot 💚. In fact, you want to try and run the algorithm yourself, so you look if the authors also happen to publish the data, the code and so on. Reproducibility is the first step to improve upon the results of an existing paper. Throughout academia as well as industry, there are several reasons why it is necessary to reproduce results:
“Scrum is like chess. You either play it as its rules state, or you don’t.”—Ken Schwaber—coinventor of Scrum
One of the most popular agile methods for Software Development is Scrum. Most developers and product managers have used it at some point in their career, it is even taught in university.
The software development process at Artory is similar to Scrum. However, during more than two years of development, we have fine-tuned the process. This post assumes basic Scrum knowledge (you can find it at ScrumGuides.org) …