Research articles

A fully convolutional neural network is used to create time-resolved three-dimensional dense segmentations of heart images. This dense motion model forms the input to a supervised system called 4Dsurvival that can efficiently predict human survival.

Neuromorphic processors promise to be a low-powered platform for deep learning, but require neural networks that are adapted for binary communication. The Whetstone method achieves this by gradually sharpening activation functions during the training process.

Deep neural networks are increasingly popular in data-intensive applications, but are power-hungry. New types of computer chips that are suited to the task of deep learning, such as memristor arrays where data handling and computing take place within the same unit, are required. A well-used deep learning model called long short-term memory, which can handle temporal sequential data analysis, is now implemented in a memristor crossbar array, promising an energy-efficient and low-footprint deep learning platform.

Not all mathematical questions can be resolved, according to Gödel’s famous incompleteness theorems. It turns out that machine learning can be vulnerable to undecidability too, as is illustrated with an example problem where learnability cannot be proved nor refuted.

Most machine learning approaches extract statistical features from data, rather than the underlying causal mechanisms. A different approach analyses information in a general way by extracting recursive patterns from data using generative models under the paradigm of computability and algorithmic information theory.

Robots need to estimate and adapt to human behaviour, especially when human dynamics change over time. Now adaptive game theory controllers can help robots adapt to human behaviour in a reaching task.