Selected Publications

[14] Incremental Missing-Data Imputation for Evolving Fuzzy Granular Prediction

Cristiano Garcia; Daniel Leite; Igor Skrjanc

We propose a modified evolving granular fuzzy rule-based model for function approximation and time series prediction in an online context where values may be missing. The fuzzy model is supplied with an incremental learning algorithm that simultaneously imputes missing data and adapts model parameters and structure over time. The evolving Fuzzy Granular Predictor (eFGP) handles single and multiple missing values on data samples by developing reduced-term consequent polynomials and utilizing time-varying granules. Missing at random (MAR) and missing completely at random (MCAR) values in nonstationary data streams are approached.

[12] Evolving fuzzy and neuro-fuzzy approaches in clustering, regression, identification, and classification: A Survey

Igor Skrjanc; Jose Iglesias; Araceli Sanchis; Daniel Leite; Edwin Lughofer; Fernando Gomide

Major assumptions in computational intelligence and machine learning consist of the availability of a historical dataset for model development, and that the resulting model will, to some extent, handle similar instances during its online operation. However, in many real-world applications, these assumptions may not hold as the amount of previously available data may be insufficient to represent the underlying system, and the environment and the system may change over time. Evolving modeling from data streams has emerged as a framework to address these issues properly by self-adaptation, single-pass learning steps and evolution as well as contraction of model components on demand and on the fly. This survey focuses on evolving fuzzy rule-based models and neuro-fuzzy networks for clustering, classification and regression and system identification in online, real-time environments where learning and model development should be performed incrementally.

[10] Evolvable fuzzy systems from data streams with missing values: With application to temporal pattern recognition and cryptocurrency prediction

Cristiano Garcia; Ahmed Esmin; Daniel Leite; Igor Skrjanc

Data streams with missing values are common in real-world applications. We present an evolving granular fuzzy-rule-based model for temporal pattern recognition and time series prediction in online nonstationary context, where values may be missing. The model has a modified rule structure that includes reduced-term consequent polynomials, and is supplied by an incremental learning algorithm that simultaneously impute missing data and update model parameters and structure. Experiments on cryptocurrency prediction show the usefulness, accuracy, processing speed, and eFGP robustness to missing values.

[8] Ensemble of evolving data clouds and fuzzy models for weather time series prediction

Eduardo Soares; Pyramo Costa; Bruno Costa; Daniel Leite

This paper describes a variation of data cloud-based intelligent method known as typicality-and-eccentricity-based method for data analysis (TEDA). The objective is to develop data-centric nonlinear and time-varying models to predict mean monthly temperature. TEDA is an incremental algorithm that considers the data density and scattering of clouds over the data space. The method does not require a priori knowledge of the dataset and user-defined parameters. However, if some knowledge about the number of clouds and rules is available, then it can be expressed through a single parameter. Past values of minimum, maximum and mean monthly temperature, as well as previous values of exogenous variables such as cloudiness, rainfall and humidity are considered in the analysis. A non-parametric Spearman correlation based method is proposed to rank and select the most relevant features and time delays for a more accurate prediction.

[6] Evolving neuro-fuzzy network for real-time high impedance fault detection and classification

Sergio Silva; Pyramo Costa; Marcio Santana; Daniel Leite

This paper concerns the application of a neuro-fuzzy learning method based on data streams for high impedance fault (HIF) detection in medium-voltage power lines. A wavelet-packet-transform-based feature extraction method combined with a variation of evolving neuro-fuzzy network with fluctuating thresholds is considered for recognition of spatial–temporal patterns in the data. Different from other statistical and intelligent approaches to the problem, the developed neuro-fuzzy model for HIF classification is not only parametrically, but also structurally adaptive to cope with nonstationarities and novelties. New neurons and connections are incrementally added to the neuro-fuzzy network when necessary for the identification of new patterns, such as faults and usual transients including sag, swell and spikes due to the switching of 3-phase capacitors and energization of transformers.

[4] Evolving Granular Fuzzy Model-Based Control of Nonlinear Dynamic Systems

Daniel Leite; Reinaldo Palhares; Victor Campos; Fernando Gomide

Unknown nonstationary processes require modeling and control design to be done in real time using streams of data collected from the process. The purpose is to stabilize the closed-loop system under changes of the operating conditions and process parameters. This paper introduces a model-based evolving granular fuzzy control approach as a step toward the development of a general framework for online modeling and control of unknown nonstationary processes with no human intervention. An incremental learning algorithm is introduced to develop and adapt the structure and parameters of the process model and controller based on information extracted from uncertain data streams. State feedback control laws and closed-loop stability are obtained from the solution of relaxed linear matrix inequalities derived from a fuzzy Lyapunov function. Bounded control inputs are also taken into account in the control system design. We explain the role of fuzzy granular data and the use of parallel distributed compensation.

[2] Evolving fuzzy granular modeling from nonstationary fuzzy data streams

Daniel Leite; Rosangela Ballini; Pyramo Costa; Fernando Gomide

As uncertain data prevail in stream applications, excessive data granularity becomes unnecessary and inefficient. This paper introduces an evolving fuzzy granular framework to learn from and model time-varying fuzzy input–output data streams. The fuzzy-set based evolving modeling framework consists of a one-pass learning algorithm capable to gradually develop the structure of rule-based models. This framework is particularly suitable to handle potentially unbounded fuzzy data streams and render singular and granular approximations of nonstationary functions. We shed light into the role of evolving fuzzy granular computing in providing high-quality approximate solutions from large volumes of real-world online data streams. An application example in weather temperature prediction using actual data is used to evaluate and illustrate the modeling approach.

[5] Comparison of Genetic and Incremental Learning Methods for Neural Network-based Electrical Machine Fault Detection

Daniel Leite

Learning of neural network structure and parameters using genetic and incremental heuristic algorithms are potential approaches to address the local optima and design issues experienced when using conventional deterministic algorithms and arbitrarily chosen network structures. This chapter presents results on the development of an evolutionary (EANN) and an evolving fuzzy granular (EGNN) neural network for detecting and classifying inter-turns short-circuit in the stator windings of induction motors. Aspects of the neural models, such as classification performance, computational complexity, and compactness, are compared with each other and with the results obtained using a conventional feedforward neural network of similar structure, but trained by a deterministic gradient descent algorithm.

[3] Evolving Linguistic Fuzzy Models from Data Streams

Daniel Leite; Fernando Gomide

This chapter outlines a new approach for online learning from imprecise data, namely, fuzzy set based evolving modeling (FBeM) approach. FBeM is an adaptive modeling framework that uses fuzzy granular objects to enclose uncertainty in the data. The FBeM algorithm is data flow driven and supports learning on an instance-per-instance recursive basis by developing and refining fuzzy models on-demand. Structurally, FBeM models combine Mamdani and functional fuzzy systems to output granular and singular approximations of nonstationary functions. Linguistic description of the behavior of the system over time is provided by information granules and associated rules. An application example on a reactive control problem, underlining the complementarity of Mamdani and functional parts of the model, shows the usefulness of the approach. More specifically, the problem concerns sensor-based robot navigation and localization, and robust obstacle avoidance.

[1] Granular Approach for Evolving Systems Modeling

Daniel Leite; Pyramo Costa; Fernando Gomide

We introduce a class of granular evolving system modeling approach within the framework of interval analysis. Our aim is to present an interval-based learning algorithm which develops both, granular and singular approximations of nonlinear nonstationary functions using singular data. The algorithm is capable of incrementally creating/adapting both model parameters and structure. Interval analysis provides rigorous bounds on approximation errors, rounding errors, and on uncertainties in data propagated during computations. The learning algorithm is simple and particularly suited to process stream of data in real time. In this paper we focus on the foundations of the approach and on the details of the learning algorithm. An application concerning economic time series forecasting illustrates the usefulness and efficiency of the approach.