Welcome to the IEEE Task Force on
Randomization-Based Neural Networks and Learning Systems

This task force is established under the Neural Networks Technical Committee (NNTC) of the Computational Intelligence Society (CIS) of the Institute of Electrical and Electronic Engineers (IEEE).

Randomization-based neural networks and learning methods have shown great potentialities, but very often they have been investigated in an undisciplined manner. For example, numerous randomization-based neural networks have been proposed independently often without mutual acquaintance and without proper performance evaluations. Similarly, there are strictly related branches of randomization-based learning methods such as the random forest variants, stochastic kernel methods and so on. All these methods have demonstrated their strengths over the years. However, their full potentials have not been realized yet. For example, deep randomization-based neural networks and deep random forests have not been well investigated.

The goal of this task force is to promote the research and applications of these methods, to demonstrate the competitive performance of randomization-based algorithms in diverse scenarios, to educate the research community about the randomization-based learning methods and their relationships, if any.

The scope of planned activities is extended but it is not limited to: randomization-based neural networks such as feedforward, convolutional, supervised, unsupervised, deep, etc.; Random Forest and its variants; randomization-based strategies in kernel methods; stochastic learning algorithms; all relevant applications.