Neural networks, a form of artificial intelligence modelled after the way our own brains process data, are the technology behind picture recognition programmes that can identify our friends on social media and wildflowers in our backyard with ease.
Neural networks can process large amounts of data quickly, but because of its design, it is hard to identify the source of mistakes that are visible to humans, such as mistaking an ankle boot for a Converse high-top. For this reason, neural networks are not as useful for more critical tasks like medical picture analysis or research. Finding those inaccuracies is now as easy as identifying mountaintops from an aeroplane thanks to a new programme created by Purdue University.