PANFIS: A Novel Incremental Learning Machine

Abstract

Most of the dynamics in real-world systems are compiled by shifts and drifts, which are uneasy to be overcome by omnipresent neuro-fuzzy systems. Nonetheless, learning in nonstationary environment entails a system owning high degree of flexibility capable of assembling its rule base autonomously according to the degree of nonlinearity contained in the system. In practice, the rule growing and pruning are carried out merely benefiting from a small snapshot of the complete training data to truncate the computational load and memory demand to the low level. An exposure of a novel algorithm, namely parsimonious network based on fuzzy inference system (PANFIS), is to this end presented herein. PANFIS can commence its learning process from scratch with an empty rule base. The fuzzy rules can be stitched up and expelled by virtue of statistical contributions of the fuzzy rules and injected datum afterward. Identical fuzzy sets may be alluded and blended to be one fuzzy set as a pursuit of a transparent rule base escalating human's interpretability. The learning and modeling performances of the proposed PANFIS are numerically validated using several benchmark problems from real-world or synthetic datasets. The validation includes comparisons with state-of-the-art evolving neuro-fuzzy methods and showcases that our new method can compete and in some cases even outperform these approaches in terms of predictive fidelity and model complexity.

Evolving Ensemble Fuzzy Classifier

“…Therefore, inconsequential rules can be located by the ERS method easily. The quadratic weight decay term is incorporated since it is capable of reducing the weight vector proportionally to its current values [47]. V. EXPERIMENTAL STUDIES We elaborate on numerical validations of pENsemble by using 15 real-world data streams and comparisons with prominent classifiers.…”

“…An evolving version of Vector quantization was designed in [41] and is algorithmic backbone of FLEXFIS [42], which was later extended to a more robust version including rule merging in [43], generalized rules and an incremental feature weighting mechanism in [44]. A generalized TSK fuzzy rule was put forward in [45]- [47] and generates a non-axis parallel ellipsoidal cluster, which happens to have better coverage and flexibility than conventional fuzzy rules [44]. Pratama et al in [47] developed the theory of rule statistical contribution borrowing the concept of hidden neuron statistical contribution in [48], [49].…”

“…A generalized TSK fuzzy rule was put forward in [45]- [47] and generates a non-axis parallel ellipsoidal cluster, which happens to have better coverage and flexibility than conventional fuzzy rules [44]. Pratama et al in [47] developed the theory of rule statistical contribution borrowing the concept of hidden neuron statistical contribution in [48], [49]. Evolving Ensemble (eEnsemble) was proposed in [14] where it makes use of eTS [7] as a base-classifier and is realised under different configurations of the ensemble classifier.…”

MSAFIS: an evolving fuzzy inference system

“…The use of evolving classifiers for activity recognition from sensor readings in ambient assisted living environments is described by [27]. In [30] and [31], two novel evolving neurofuzzy algorithms are developed. A new concept is addressed in [41] for handling drifts in data streams during the run of online evolving modeling processes in a regression context.…”

Online active supervision of an evolving classifier for customized-gesture-command learning

“…PANFIS [48] can split, merge or remove fuzzy rules to improve concept drifts adaptation. pClass [49] can also recall old rules to better follow reoccurring concepts.…”