Researchers from Microsoft Research Cambridge are developing software to help computers learn from experience, creating the next generation of machine intelligence.
Scientists have worked for decades to try to create intelligence in computers. Traditional approaches relied on hand-crafted solutions, and had limited applicability. New techniques being developed at Microsoft Research Cambridge, and at other institutions around the world, are based on computers which can learn for themselves by analysing large sets of data.
The key to learning from examples is to recognise that real-world data is full of complexity, ambiguity, and uncertainty. Computers can be programmed to handle these challenges by using probabilities.
“By embracing the mathematics of uncertainty, we are able to create machines that can learn from data, and to apply these to many real-world applications ranging from image editing, to highly interactive games” says Professor Christopher Bishop at Microsoft Research Cambridge.
Visitors to the exhibit will be able to interact with 3-D medical images and see how machine intelligence techniques can highlight lesions automatically. They will also be able to test the ability of ‘LiveObject’ to recognise everyday objects such as pens and mobile phones, and will be able to explore the science of uncertainty for themselves through a variety of hands-on demonstrations.
Exhibited by Microsoft Research Cambridge.
See all exhibits from 2010