Qatar University

Nokia Corporation

Moncef Gabbouj

Serkan Kiranyaz

Turker Ince

Imed Bouazizi

Mehdi Rezaei

Alexandros Iosifidis

Certain embodiments may generally relate to various techniques for machine learning. Feed-forward, fully-connected Artificial Neural Networks (ANNs), or the so-called Multi-Layer Perceptrons (MLPs) are well-known universal approximators. However, their learning performance may vary significantly depending on the function or the solution space that they attempt to approximate for learning. This is because they are based on a loose and crude model of the biological neurons promising only a linear transformation followed by a nonlinear activation function. Therefore, while they learn very well those problems with a monotonous, relatively simple and linearly separable solution space, they may entirely fail to do so when the solution space is highly nonlinear and complex. In order to address this drawback and also to accomplish a more generalized model of biological neurons and learning systems, Generalized Operational Perceptrons (GOPs) may be formed and they may encapsulate many linear and nonlinear operators.

Generalized operational perceptrons: new generation artificial neural networks

A method of detecting abnormal heartbeats includes providing a library of abnormal beat synthesis (ABS) filters, wherein each ABS filter corresponds to a specific cause of a cardiac problem. The method further includes obtaining an ECG of a normal heartbeat of a person and applying an ABS filter from the library of ABS filters to the ECG of the normal heartbeat of the person to generate a potential abnormal ECG. The method further includes monitoring a heartbeat of the person and classifying each heartbeat as either normal or abnormal.

Personalized ECG monitoring for early detection of cardiac abnormalities

In accordance with an aspect of the invention, flexible burst boundaries, within an allowable range, allow statistical multiplexing to be performed in conjunction with time slicing in a digital broadcast network. The signalling of the delta-t is performed such that data of the service is not missed. Similar to deterministic multiplexing, aspects of the invention split the multiplex into several time-cycles and allocate, for each service, a time slot (data burst) from the total time-cycle according to the service's average bit rate. In accordance with aspects of the invention, however, the boundaries of each data burst are allowed to grow or shrink by a limited amount.

Growing community of inventors

Tampere, Finland

Moncef Gabbouj