Abstract:The Metacognitive Scaffolding Learning Machine (McSLM), combining the concept of metacognition—what-to-learn, how-to-learn, and when-to-learn, and the Scaffolding theory—a tutoring theory for a learner to learn a complex task, has been successfully developed to enhance the capability of Evolving Intelligent Systems (EIS) in processing non-stationary data streams. Three issues, namely uncertainty, temporal behaviour, unknown system order, are however uncharted by any existing McSLMs and all McSLMs in the literature are designed for classification problems. This paper proposes a novel McSLM, called Recurrent Interval-Valued Metacognitive Scaffolding Fuzzy Neural Network (RIVMcSFNN) and used to solve regression and time-series modelling problems from data streams. RIVMcSFNN presents a novel recurrent network architecture as a cognitive constituent, which features double local recurrent connections at both the hidden layer and the consequent layer. The new recurrent network architecture is driven by the interval-valued multivariate Gaussian function in the hidden node and the nonlinear Wavelet function in the consequent node. As with its predecessors, the RIVMcSFNN characterises an open structure, where it can automatically grow, prune, adjust, merge, recall its hidden node and can select relevant data samples on the fly using an online active learning methodology. The RIVMcSFNN is also equipped with the online dimensionality reduction technique to cope with the curse of dimensionality. All learning mechanisms are carried out in the single-pass and local learning mode and actualise the plug-and-play learning principle, which aims to minimise the use of pre-and/or post-training steps. The efficacy of our algorithm was tested using numerous data-driven modelling problems and comprehensive comparisons with its counterparts. The RIVMcSFNN demonstrated substantial improvements in both accuracy and complexity against existing variants of the McSLMs and EISs.

Meta-cognitive Fuzzy Extreme Learning Machine.

Extreme Learning Machine with Multiple Kernels

A Fast Incremental Method Based on Regularized Extreme Learning Machine

Dolphin Swarm Extreme Learning Machine.

Fuzzy Extreme Learning Machine for A Class of Fuzzy Inference Systems

An Incremental Type-2 Meta-Cognitive Extreme Learning Machine.

Memetic Extreme Learning Machine.

Sequential active learning using meta-cognitive extreme learning machine.

Data Driven Modelling Based on Recurrent Interval-Valued Metacognitive Scaffolding Fuzzy Neural Network.

Online Regularized and Kernelized Extreme Learning Machines with Forgetting Mechanism

Semi-supervised Multi-kernel Extreme Learning Machine.

Fuzzy Extreme Learning Machine for Classification

Adaptive Online Sequential Extreme Learning Machine for Dynamic Modeling

Online sequential extreme learning machine with forgetting mechanism.

A Novel Extreme Learning Machine Based on Hybrid Kernel Function.

A Memetic Algorithm Based Extreme Learning Machine for Classification

An Adaptive Extreme Learning Machine Algorithm and Its Application on Face Recognition

A Kind of Parameters Self-adjusting Extreme Learning Machine

Fuzzy ELM for classification based on feature space

An extreme learning machine model based on adaptive multi-fusion chaotic sparrow search algorithm for regression and classification

Extreme Learning Machine: A New Learning Scheme of Feedforward Neural Networks