scientific data

One application of AI, and in particular machine learning, is to process and interpret scientific data. This can often be voluminous, indeed the World Wide Web was originally developed to help share the vast amunts of data generated at CERN, it is thus an early example of big data. Some forms of processing may simply be to reduce the dataset, ignoring common events or patterns and bringing more interesting ones to the scientists' attention. However, forms of rule induction can be used to infer scientific laws directly from data, in a similar way to how Keplar spoted the relationships between planets' orbital distance and speed, namely that they "sweep out equal areas in equal times". Automated systems are able to try many different forms of equations and then choose ones that are both simple and fit the data well.

Used on page 391