Lifelong Generative Adversarial Autoencoder

Lifelong learning describes an ability that enables humans to continually acquire and learn new information without forgetting. This capability, common to humans and animals, has lately been identified as an essential function for an artificial intelligence system aiming to learn continuously from a stream of data during a certain period of time. However, modern neural networks suffer from degenerated performance when learning multiple domains sequentially and fail to recognize past learned tasks after being retrained. This corresponds to catastrophic forgetting and is ultimately induced by replacing the parameters associated with previously learned tasks with new values. One approach in lifelong learning is the generative replay mechanism (GRM) that trains a powerful generator as the generative replay network, implemented by a variational autoencoder (VAE) or a generative adversarial network (GAN). In this article, we study the forgetting behavior of GRM-based learning systems by developing a new theoretical framework in which the forgetting process is expressed as an increase in the model's risk during the training. Although many recent attempts have provided high-quality generative replay samples by using GANs, they are limited to mainly downstream tasks due to the lack of inference. Inspired by the theoretical analysis while aiming to address the drawbacks of existing approaches, we propose the lifelong generative adversarial autoencoder (LGAA). LGAA consists of a generative replay network and three inference models, each addressing the inference of a different type of latent variable. The experimental results show that LGAA learns novel visual concepts without forgetting and can be applied to a wide range of downstream tasks.

Federated Learning Empowered Real-Time Medical Data Processing Method for Smart Healthcare

Computer-aided diagnosis (CAD) has always been an important research topic for applying artificial intelligence in smart healthcare. Sufficient medical data are one of the most critical factors in CAD research. However, medical data are usually obtained in chronological order and cannot be collected all at once, which poses difficulties for the application of deep learning technology in the medical field. The traditional batch learning method consumes considerable time and space resources for real-time medical data, and the incremental learning method often leads to catastrophic forgetting. To solve these problems, we propose a real-time medical data processing method based on federated learning. We divide the process into the model stage and the exemplar stage. In the model stage, we use the federated learning method to fuse the old and new models to mitigate the catastrophic forgetting problem of the new model. In the exemplar stage, we use the most representative exemplars selected from the old data to help the new model review the old knowledge, which further mitigates the catastrophic forgetting problem of the new model. We use this method to conduct experiments on a simulated medical real-time data stream. The experimental results show that our method can learn a disease diagnosis model from a continuous medical real-time data stream. As the amount of data increases, the performance of the disease diagnosis model continues to improve, and the catastrophic forgetting problem has been effectively mitigated. Compared with the traditional batch learning method, our method can significantly save time and space resources.

A comprehensive AI model development framework for consistent Gleason grading

BACKGROUND

Artificial Intelligence(AI)-based solutions for Gleason grading hold promise for pathologists, while image quality inconsistency, continuous data integration needs, and limited generalizability hinder their adoption and scalability.

METHODS

We present a comprehensive digital pathology workflow for AI-assisted Gleason grading. It incorporates A!MagQC (image quality control), A!HistoClouds (cloud-based annotation), Pathologist-AI Interaction (PAI) for continuous model improvement, Trained on Akoya-scanned images only, the model utilizes color augmentation and image appearance migration to address scanner variations. We evaluate it on Whole Slide Images (WSI) from another five scanners and conduct validations with pathologists to assess AI efficacy and PAI.

RESULTS

Our model achieves an average F1 score of 0.80 on annotations and 0.71 Quadratic Weighted Kappa on WSIs for Akoya-scanned images. Applying our generalization solution increases the average F1 score for Gleason pattern detection from 0.73 to 0.88 on images from other scanners. The model accelerates Gleason scoring time by 43% while maintaining accuracy. Additionally, PAI improve annotation efficiency by 2.5 times and led to further improvements in model performance.

CONCLUSIONS

This pipeline represents a notable advancement in AI-assisted Gleason grading for improved consistency, accuracy, and efficiency. Unlike previous methods limited by scanner specificity, our model achieves outstanding performance across diverse scanners. This improvement paves the way for its seamless integration into clinical workflows.

A Dirichlet Process Mixture of Robust Task Models for Scalable Lifelong Reinforcement Learning

While reinforcement learning (RL) algorithms are achieving state-of-the-art performance in various challenging tasks, they can easily encounter catastrophic forgetting or interference when faced with lifelong streaming information. In this article, we propose a scalable lifelong RL method that dynamically expands the network capacity to accommodate new knowledge while preventing past memories from being perturbed. We use a Dirichlet process mixture to model the nonstationary task distribution, which captures task relatedness by estimating the likelihood of task-to-cluster assignments and clusters the task models in a latent space. We formulate the prior distribution of the mixture as a Chinese restaurant process (CRP) that instantiates new mixture components as needed. The update and expansion of the mixture are governed by the Bayesian nonparametric framework with an expectation maximization (EM) procedure, which dynamically adapts the model complexity without explicit task boundaries or heuristics. Moreover, we use the domain randomization technique to train robust prior parameters for the initialization of each task model in the mixture; thus, the resulting model can better generalize and adapt to unseen tasks. With extensive experiments conducted on robot navigation and locomotion domains, we show that our method successfully facilitates scalable lifelong RL and outperforms relevant existing methods.

Model Behavior Preserving for Class-Incremental Learning

Deep models have shown to be vulnerable to catastrophic forgetting, a phenomenon that the recognition performance on old data degrades when a pre-trained model is fine-tuned on new data. Knowledge distillation (KD) is a popular incremental approach to alleviate catastrophic forgetting. However, it usually fixes the absolute values of neural responses for isolated historical instances, without considering the intrinsic structure of the responses by a convolutional neural network (CNN) model. To overcome this limitation, we recognize the importance of the global property of the whole instance set and treat it as a behavior characteristic of a CNN model relevant to model incremental learning. On this basis: 1) we design an instance neighborhood-preserving (INP) loss to maintain the order of pair-wise instance similarities of the old model in the feature space; 2) we devise a label priority-preserving (LPP) loss to preserve the label ranking lists within instance-wise label probability vectors in the output space; and 3) we introduce an efficient derivable ranking algorithm for calculating the two loss functions. Extensive experiments conducted on CIFAR100 and ImageNet show that our approach achieves the state-of-the-art performance.

Few Shot Class Incremental Learning via Efficient Prototype Replay and Calibration

Few shot class incremental learning (FSCIL) is an extremely challenging but valuable problem in real-world applications. When faced with novel few shot tasks in each incremental stage, it should take into account both catastrophic forgetting of old knowledge and overfitting of new categories with limited training data. In this paper, we propose an efficient prototype replay and calibration (EPRC) method with three stages to improve classification performance. We first perform effective pre-training with rotation and mix-up augmentations in order to obtain a strong backbone. Then a series of pseudo few shot tasks are sampled to perform meta-training, which enhances the generalization ability of both the feature extractor and projection layer and then helps mitigate the over-fitting problem of few shot learning. Furthermore, an even nonlinear transformation function is incorporated into the similarity computation to implicitly calibrate the generated prototypes of different categories and alleviate correlations among them. Finally, we replay the stored prototypes to relieve catastrophic forgetting and rectify prototypes to be more discriminative in the incremental-training stage via an explicit regularization within the loss function. The experimental results on CIFAR-100 and miniImageNet demonstrate that our EPRC significantly boosts the classification performance compared with existing mainstream FSCIL methods.

Dynamic Self-Supervised Teacher-Student Network Learning

Lifelong learning (LLL) represents the ability of an artificial intelligence system to learn successively a sequence of different databases. In this paper we introduce the Dynamic Self-Supervised Teacher-Student Network (D-TS), representing a more general LLL framework, where the Teacher is implemented as a dynamically expanding mixture model which automatically increases its capacity to deal with a growing number of tasks. We propose the Knowledge Discrepancy Score (KDS) criterion for measuring the relevance of the incoming information characterizing a new task when compared to the existing knowledge accumulated by the Teacher module from its previous training. The KDS ensures a light Teacher architecture while also enabling to reuse the learned knowledge whenever appropriate, accelerating the learning of given tasks. The Student module is implemented as a lightweight probabilistic generative model. We introduce a novel self-supervised learning procedure for the Student that allows to capture cross-domain latent representations from the entire knowledge accumulated by the Teacher as well as from novel data. We perform several experiments which show that D-TS can achieve the state of the art results in LLL while requiring fewer parameters than other methods.

Enhanced rainfall prediction performance via hybrid empirical-singular-wavelet-fuzzy approaches

Rainfall is a vital process in the hydrological cycle of the globe. Accessing reliable and accurate rainfall data is crucial for water resources operation, flood control, drought warning, irrigation, and drainage. In the present study, the main objective is to develop a predictive model to enhance daily rainfall prediction accuracy with an extended time horizon. In the literature, various methods for the prediction of daily rainfall data for short lead times are presented. However, due to the complex and random nature of rainfall, in general, they yield inaccurate prediction results. Generically, rainfall predictive models require many physical meteorological variables and consist of challenging mathematical processes that require high computational power. Furthermore, due to the nonlinear and chaotic nature of rainfall, observed raw data typically has to be decomposed into its trend cycle, seasonality, and stochastic components before being fed into the predictive model. The present study proposes a novel singular spectrum analysis (SSA)-based approach for decomposing observed raw data into its hierarchically energetic pertinent features. To this end, in addition to the stand-alone fuzzy logic model, preprocessing methods SSA, empirical mode decomposition (EMD), and commonly used discrete wavelet transform (DWT) are incorporated into the fuzzy models which are named as hybrid SSA-fuzzy, EMD-fuzzy, W-fuzzy models, respectively. In this study, fuzzy, hybrid SSA-fuzzy, EMD-fuzzy, and W-fuzzy models are developed to enhance the daily rainfall prediction accuracy and improve the prediction time span up to 3 days via three (3) stations' data in Turkey. The proposed SSA-fuzzy model is compared with fuzzy, hybrid EMD-fuzzy, and widely used hybrid W-fuzzy models in predicting daily rainfall in three distinctive locations up to a 3-day time horizon. Improved accuracy in predicting daily rainfall is provided by the SSA-fuzzy, W-fuzzy, and EMD-fuzzy models compared to the stand-alone fuzzy model based on mean square error (MSE) and the Nash-Sutcliffe coefficient of efficiency (CE) model assessment metrics. Specifically, the advocated SSA-fuzzy model is found to be superior in accuracy to hybrid EMD-fuzzy and W-fuzzy models in predicting daily rainfall for all time spans. The results reveal that, with its easy-to-use features, the advocated SSA-fuzzy modeling tool in this study is a promising principled method for its possible future implementations not only in hydrological studies but in water resources and hydraulics engineering and all scientific disciplines where future state space prediction of a vague nature and stochastic dynamical system is important.

Lifelong Mixture of Variational Autoencoders

In this article, we propose an end-to-end lifelong learning mixture of experts. Each expert is implemented by a variational autoencoder (VAE). The experts in the mixture system are jointly trained by maximizing a mixture of individual component evidence lower bounds (MELBO) on the log-likelihood of the given training samples. The mixing coefficients in the mixture model control the contributions of each expert in the global representation. These are sampled from a Dirichlet distribution whose parameters are determined through nonparametric estimation during lifelong learning. The model can learn new tasks fast when these are similar to those previously learned. The proposed lifelong mixture of VAE (L-MVAE) expands its architecture with new components when learning a completely new task. After the training, our model can automatically determine the relevant expert to be used when fed with new data samples. This mechanism benefits both the memory efficiency and the required computational cost as only one expert is used during the inference. The L-MVAE inference model is able to perform interpolations in the joint latent space across the data domains associated with different tasks and is shown to be efficient for disentangled learning representation.

Autonomous learning for fuzzy systems: a review

As one of the three pillars in computational intelligence, fuzzy systems are a powerful mathematical tool widely used for modelling nonlinear problems with uncertainties. Fuzzy systems take the form of linguistic IF-THEN fuzzy rules that are easy to understand for human. In this sense, fuzzy inference mechanisms have been developed to mimic human reasoning and decision-making. From a data analytic perspective, fuzzy systems provide an effective solution to build precise predictive models from imprecise data with great transparency and interpretability, thus facilitating a wide range of real-world applications. This paper presents a systematic review of modern methods for autonomously learning fuzzy systems from data, with an emphasis on the structure and parameter learning schemes of mainstream evolving, evolutionary, reinforcement learning-based fuzzy systems. The main purpose of this paper is to introduce the underlying concepts, underpinning methodologies, as well as outstanding performances of the state-of-the-art methods. It serves as a one-stop guide for readers learning the representative methodologies and foundations of fuzzy systems or who desire to apply fuzzy-based autonomous learning in other scientific disciplines and applied fields.

Re-Training of Convolutional Neural Networks for Glottis Segmentation in Endoscopic High-Speed Videos

Endoscopic high-speed video (HSV) systems for visualization and assessment of vocal fold dynamics in the larynx are diverse and technically advancing. To consider resulting "concepts shifts" for neural network (NN)-based image processing, re-training of already trained and used NNs is necessary to allow for sufficiently accurate image processing for new recording modalities. We propose and discuss several re-training approaches for convolutional neural networks (CNN) being used for HSV image segmentation. Our baseline CNN was trained on the BAGLS data set (58,750 images). The new BAGLS-RT data set consists of additional 21,050 images from previously unused HSV systems, light sources, and different spatial resolutions. Results showed that increasing data diversity by means of preprocessing already improves the segmentation accuracy (mIoU + 6.35%). Subsequent re-training further increases segmentation performance (mIoU + 2.81%). For re-training, finetuning with dynamic knowledge distillation showed the most promising results. Data variety for training and additional re-training is a helpful tool to boost HSV image segmentation quality. However, when performing re-training, the phenomenon of catastrophic forgetting should be kept in mind, i.e., adaption to new data while forgetting already learned knowledge.

Lifelong Teacher-Student Network Learning

A unique cognitive capability of humans consists in their ability to acquire new knowledge and skills from a sequence of experiences. Meanwhile, artificial intelligence systems are good at learning only the last given task without being able to remember the databases learnt in the past. We propose a novel lifelong learning methodology by employing a Teacher-Student network framework. While the Student module is trained with a new given database, the Teacher module would remind the Student about the information learnt in the past. The Teacher, implemented by a Generative Adversarial Network (GAN), is trained to preserve and replay past knowledge corresponding to the probabilistic representations of previously learnt databases. Meanwhile, the Student module is implemented by a Variational Autoencoder (VAE) which infers its latent variable representation from both the output of the Teacher module as well as from the newly available database. Moreover, the Student module is trained to capture both continuous and discrete underlying data representations across different domains. The proposed lifelong learning framework is applied in supervised, semi-supervised and unsupervised training.

Lifelong Adaptive Machine Learning for Sensor-Based Human Activity Recognition Using Prototypical Networks

Continual learning (CL), also known as lifelong learning, is an emerging research topic that has been attracting increasing interest in the field of machine learning. With human activity recognition (HAR) playing a key role in enabling numerous real-world applications, an essential step towards the long-term deployment of such systems is to extend the activity model to dynamically adapt to changes in people's everyday behavior. Current research in CL applied to the HAR domain is still under-explored with researchers exploring existing methods developed for computer vision in HAR. Moreover, analysis has so far focused on task-incremental or class-incremental learning paradigms where task boundaries are known. This impedes the applicability of such methods for real-world systems. To push this field forward, we build on recent advances in the area of continual learning and design a lifelong adaptive learning framework using Prototypical Networks, LAPNet-HAR, that processes sensor-based data streams in a task-free data-incremental fashion and mitigates catastrophic forgetting using experience replay and continual prototype adaptation. Online learning is further facilitated using contrastive loss to enforce inter-class separation. LAPNet-HAR is evaluated on five publicly available activity datasets in terms of its ability to acquire new information while preserving previous knowledge. Our extensive empirical results demonstrate the effectiveness of LAPNet-HAR in task-free CL and uncover useful insights for future challenges.

Online Reinforcement Learning Control by Direct Heuristic Dynamic Programming: From Time-Driven to Event-Driven

In this work, time-driven learning refers to the machine learning method that updates parameters in a prediction model continuously as new data arrives. Among existing approximate dynamic programming (ADP) and reinforcement learning (RL) algorithms, the direct heuristic dynamic programming (dHDP) has been shown an effective tool as demonstrated in solving several complex learning control problems. It continuously updates the control policy and the critic as system states continuously evolve. It is therefore desirable to prevent the time-driven dHDP from updating due to insignificant system event such as noise. Toward this goal, we propose a new event-driven dHDP. By constructing a Lyapunov function candidate, we prove the uniformly ultimately boundedness (UUB) of the system states and the weights in the critic and the control policy networks. Consequently, we show the approximate control and cost-to-go function approaching Bellman optimality within a finite bound. We also illustrate how the event-driven dHDP algorithm works in comparison to the original time-driven dHDP.

Incremental learning for detection in X-ray luggage perspective images

Convolutional neural networks have achieved remarkable results in the detection of X-ray luggage contraband. However, with an increase in contraband classes and substantial artificial transformation, the offline network training method has been unable to accurately detect the rapidly growing new classes of contraband. The current model cannot incrementally learn the newly appearing classes in real time without retraining the model. When the quantity of different types of contraband is not evenly distributed in the real-time detection process, the convolution neural network that is optimized by the gradient descent method will produce catastrophic forgetting, which means learning new knowledge and forgetting old knowledge, and the detection effect on the old classes will suddenly decline. To overcome this problem, this paper proposes an incremental learning method for online continuous learning of models and incrementally learns and detects new classes in the absence of old classes in the new classes. First, we perform parameter compression on the original network by distillation to ensure stable identification of the old classes. Second, the area proposal subnetwork and object detection subnetwork are incrementally learned to obtain the recognition ability of the new classes. In addition, this paper designs a new loss function, which causes the network to avoid catastrophic forgetting and stably detect the object of the new contraband classes. To reliably verify the model, this paper produces a multi-angle dataset for security perspective images. A total of 10 classes of contraband are tested, and the interference between two object detections is analyzed by model parameters. The experimental results show that the model can stably perform new contraband object learning even when there is an uneven distribution of data types.

R. N, B. R. S, Lahza H, Lahza HFM. A survey on detecting healthcare concept drift in AI/ML models from a finance perspective

Data is incredibly significant in today's digital age because data represents facts and numbers from our regular life transactions. Data is no longer arriving in a static form; it is now arriving in a streaming fashion. Data streams are the arrival of limitless, continuous, and rapid data. The healthcare industry is a major generator of data streams. Processing data streams is extremely complex due to factors such as volume, pace, and variety. Data stream classification is difficult owing to idea drift. Concept drift occurs in supervised learning when the statistical properties of the target variable that the model predicts change unexpectedly. We focused on solving various forms of concept drift problems in healthcare data streams in this research, and we outlined the existing statistical and machine learning methodologies for dealing with concept drift. It also emphasizes the use of deep learning algorithms for concept drift detection and describes the various healthcare datasets utilized for concept drift detection in data stream categorization.

A Taxonomy of Food Supply Chain Problems from a Computational Intelligence Perspective

In the last few years, the Internet of Things, and other enabling technologies, have been progressively used for digitizing Food Supply Chains (FSC). These and other digitalization-enabling technologies are generating a massive amount of data with enormous potential to manage supply chains more efficiently and sustainably. Nevertheless, the intricate patterns and complexity embedded in large volumes of data present a challenge for systematic human expert analysis. In such a data-driven context, Computational Intelligence (CI) has achieved significant momentum to analyze, mine, and extract the underlying data information, or solve complex optimization problems, striking a balance between productive efficiency and sustainability of food supply systems. Although some recent studies have sorted the CI literature in this field, they are mainly oriented towards a single family of CI methods (a group of methods that share common characteristics) and review their application in specific FSC stages. As such, there is a gap in identifying and classifying FSC problems from a broader perspective, encompassing the various families of CI methods that can be applied in different stages (from production to retailing) and identifying the problems that arise in these stages from a CI perspective. This paper presents a new and comprehensive taxonomy of FSC problems (associated with agriculture, fish farming, and livestock) from a CI approach; that is, it defines FSC problems (from production to retail) and categorizes them based on how they can be modeled from a CI point of view. Furthermore, we review the CI approaches that are more commonly used in each stage of the FSC and in their corresponding categories of problems. We also introduce a set of guidelines to help FSC researchers and practitioners to decide on suitable families of methods when addressing any particular problems they might encounter. Finally, based on the proposed taxonomy, we identify and discuss challenges and research opportunities that the community should explore to enhance the contributions that CI can bring to the digitization of the FSC.

An overview of complex data stream ensemble classification

With the rapid development of information technology, data streams in various fields are showing the characteristics of rapid arrival, complex structure and timely processing. Complex types of data streams make the classification performance worse. However, ensemble classification has become one of the main methods of processing data streams. Ensemble classification performance is better than traditional single classifiers. This article introduces the ensemble classification algorithms of complex data streams for the first time. Then overview analyzes the advantages and disadvantages of these algorithms for steady-state, concept drift, imbalanced, multi-label and multi-instance data streams. At the same time, the application fields of data streams are also introduced which summarizes the ensemble algorithms processing text, graph and big data streams. Moreover, it comprehensively summarizes the verification technology, evaluation indicators and open source platforms of complex data streams mining algorithms. Finally, the challenges and future research directions of ensemble learning algorithms dealing with uncertain, multi-type, delayed, multi-type concept drift data streams are given.

Continual Learning Using Bayesian Neural Networks

Continual learning models allow them to learn and adapt to new changes and tasks over time. However, in continual and sequential learning scenarios, in which the models are trained using different data with various distributions, neural networks (NNs) tend to forget the previously learned knowledge. This phenomenon is often referred to as catastrophic forgetting. The catastrophic forgetting is an inevitable problem in continual learning models for dynamic environments. To address this issue, we propose a method, called continual Bayesian learning networks (CBLNs), which enables the networks to allocate additional resources to adapt to new tasks without forgetting the previously learned tasks. Using a Bayesian NN, CBLN maintains a mixture of Gaussian posterior distributions that are associated with different tasks. The proposed method tries to optimize the number of resources that are needed to learn each task and avoids an exponential increase in the number of resources that are involved in learning multiple tasks. The proposed method does not need to access the past training data and can choose suitable weights to classify the data points during the test time automatically based on an uncertainty criterion. We have evaluated the method on the MNIST and UCR time-series data sets. The evaluation results show that the method can address the catastrophic forgetting problem at a promising rate compared to the state-of-the-art models.

Online sequential ensembling of predictive fuzzy systems

Evolving fuzzy systems (EFS) have enjoyed a wide attraction in the community to handle learning from data streams in an incremental, single-pass and transparent manner. The main concentration so far lied in the development of approaches for single EFS models, basically used for prediction purposes. Forgetting mechanisms have been used to increase their flexibility, especially for the purpose to adapt quickly to changing situations such as drifting data distributions. These require forgetting factors steering the degree of timely out-weighing older learned concepts, whose adequate setting in advance or in adaptive fashion is not an easy and not a fully resolved task. In this paper, we propose a new concept of learning fuzzy systems from data streams, which we call online sequential ensembling of fuzzy systems (OS-FS). It is able to model the recent dependencies in streams on a chunk-wise basis: for each new incoming chunk, a new fuzzy model is trained from scratch and added to the ensemble (of fuzzy systems trained before). This induces (i) maximal flexibility in terms of being able to apply variable chunk sizes according to the actual system delay in receiving target values and (ii) fast reaction possibilities in the case of arising drifts. The latter are realized with specific prediction techniques on new data chunks based on the sequential ensemble members trained so far over time. We propose four different prediction variants including various weighting concepts in order to put higher weights on the members with higher inference certainty during the amalgamation of predictions of single members to a final prediction. In this sense, older members, which keep in mind knowledge about past states, may get dynamically reactivated in the case of cyclic drifts, which induce dynamic changes in the process behavior which are re-occurring from time to time later. Furthermore, we integrate a concept for properly resolving possible contradictions among members with similar inference certainties. The reaction onto drifts is thus autonomously handled on demand and on the fly during the prediction stage (and not during model adaptation/evolution stage as conventionally done in single EFS models), which yields enormous flexibility. Finally, in order to cope with large-scale and (theoretically) infinite data streams within a reasonable amount of prediction time, we demonstrate two concepts for pruning past ensemble members, one based on atypical high error trends of single members and one based on the non-diversity of ensemble members. The results based on two data streams showed significantly improved performance compared to single EFS models in terms of a better convergence of the accumulated chunk-wise ahead prediction error trends, especially in the case of regular and cyclic drifts. Moreover, the more advanced prediction schemes could significantly outperform standard averaging over all members’ outputs. Furthermore, resolving contradictory outputs among members helped to improve the performance of the sequential ensemble further. Results on a wider range of data streams from different application scenarios showed (i) improved error trend lines over single EFS models, as well as over related AI methods OS-ELM and MLPs neural networks retrained on data chunks, and (ii) slightly worse trend lines than on-line bagged EFS (as specific EFS ensembles), but with around 100 times faster processing times (achieving low processing times way below requiring milli-seconds for single samples updates).

Cost-Sensitive Classification for Evolving Data Streams with Concept Drift and Class Imbalance

Class imbalance and concept drift are two primary principles that exist concurrently in data stream classification. Although the two issues have drawn enough attention separately, the joint treatment largely remains unexplored. Moreover, the class imbalance issue is further complicated if data streams with concept drift. A novel Cost-Sensitive based Data Stream (CSDS) classification is introduced to overcome the two issues simultaneously. The CSDS considers cost information during the procedures of data preprocessing and classification. During the data preprocessing, a cost-sensitive learning strategy is introduced into the ReliefF algorithm for alleviating the class imbalance at the data level. In the classification process, a cost-sensitive weighting schema is devised to enhance the overall performance of the ensemble. Besides, a change detection mechanism is embedded in our algorithm, which guarantees that an ensemble can capture and react to drift promptly. Experimental results validate that our method can obtain better classification results under different imbalanced concept drifting data stream scenarios.

Modeling and augmenting of fMRI data using deep recurrent variational auto-encoder

Objective. Recently, deep learning models have been successfully applied in functional magnetic resonance imaging (fMRI) modeling and associated applications. However, there still exist at least two challenges. Firstly, due to the lack of sufficient data, deep learning models tend to suffer from overfitting in the training process. Secondly, it is still challenging to model the temporal dynamics from fMRI, due to that the brain state is continuously changing over scan time. In addition, existing methods rarely studied and applied fMRI data augmentation.Approach. In this work, we construct a deep recurrent variational auto-encoder (DRVAE) that combined variational auto-encoder and recurrent neural network, aiming to address all of the above mentioned challenges. The encoder of DRVAE can extract more generalized temporal features from assumed Gaussian distribution of input data, and the decoder of DRVAE can generate new data to increase training samples and thus partially relieve the overfitting issue. The recurrent layers in DRVAE are designed to effectively model the temporal dynamics of functional brain activities. LASSO (least absolute shrinkage and selection operator) regression is applied on the temporal features and input fMRI data to estimate the corresponding spatial networks.Main results. Extensive experimental results on seven tasks from HCP dataset showed that the DRVAE and LASSO framework can learn meaningful temporal patterns and spatial networks from both real data and generated data. The results on group-wise data and single subject suggest that the brain activities may follow certain distribution. Moreover, we applied DRVAE on four resting state fMRI datasets from ADHD-200 for data augmentation, and the results showed that the classification performances on augmented datasets have been considerably improved.Significance. The proposed method can not only derive meaningful temporal features and spatial networks from fMRI, but also generate high-quality new data for fMRI data augmentation and associated applications.

Concept Drift Detection via Equal Intensity k-Means Space Partitioning

The data stream poses additional challenges to statistical classification tasks because distributions of the training and target samples may differ as time passes. Such a distribution change in streaming data is called concept drift. Numerous histogram-based distribution change detection methods have been proposed to detect drift. Most histograms are developed on the grid-based or tree-based space partitioning algorithms which makes the space partitions arbitrary, unexplainable, and may cause drift blind spots. There is a need to improve the drift detection accuracy for the histogram-based methods with the unsupervised setting. To address this problem, we propose a cluster-based histogram, called equal intensity k -means space partitioning (EI-kMeans). In addition, a heuristic method to improve the sensitivity of drift detection is introduced. The fundamental idea of improving the sensitivity is to minimize the risk of creating partitions in distribution offset regions. Pearson's chi-square test is used as the statistical hypothesis test so that the test statistics remain independent of the sample distribution. The number of bins and their shapes, which strongly influence the ability to detect drift, are determined dynamically from the sample based on an asymptotic constraint in the chi-square test. Accordingly, three algorithms are developed to implement concept drift detection, including a greedy centroids initialization algorithm, a cluster amplify-shrink algorithm, and a drift detection algorithm. For drift adaptation, we recommend retraining the learner if a drift is detected. The results of experiments on the synthetic and real-world datasets demonstrate the advantages of EI-kMeans and show its efficacy in detecting concept drift.

Incremental Learning in Modelling Process Analysis Technology (PAT)-An Important Tool in the Measuring and Control Circuit on the Way to the Smart Factory

To meet the demands of the chemical and pharmaceutical process industry for a combination of high measurement accuracy, product selectivity, and low cost of ownership, the existing measurement and evaluation methods have to be further developed. This paper demonstrates the attempt to combine future Raman photometers with promising evaluation methods. As part of the investigations presented here, a new and easy-to-use evaluation method based on a self-learning algorithm is presented. This method can be applied to various measurement methods and is carried out here using an example of a Raman spectrometer system and an alcohol-water mixture as demonstration fluid. The spectra’s chosen bands can be later transformed to low priced and even more robust Raman photometers. The evaluation method gives more precise results than the evaluation through classical methods like one primarily used in the software package Unscrambler. This technique increases the accuracy of detection and proves the concept of Raman process monitoring for determining concentrations. In the example of alcohol/water, the computation time is less, and it can be applied to continuous column monitoring.

Classification of the drifting data streams using heterogeneous diversified dynamic class-weighted ensemble

Data streams can be defined as the continuous stream of data coming from different sources and in different forms. Streams are often very dynamic, and its underlying structure usually changes over time, which may result to a phenomenon called concept drift. When solving predictive problems using the streaming data, traditional machine learning models trained on historical data may become invalid when such changes occur. Adaptive models equipped with mechanisms to reflect the changes in the data proved to be suitable to handle drifting streams. Adaptive ensemble models represent a popular group of these methods used in classification of drifting data streams. In this paper, we present the heterogeneous adaptive ensemble model for the data streams classification, which utilizes the dynamic class weighting scheme and a mechanism to maintain the diversity of the ensemble members. Our main objective was to design a model consisting of a heterogeneous group of base learners (Naive Bayes, k-NN, Decision trees), with adaptive mechanism which besides the performance of the members also takes into an account the diversity of the ensemble. The model was experimentally evaluated on both real-world and synthetic datasets. We compared the presented model with other existing adaptive ensemble methods, both from the perspective of predictive performance and computational resource requirements.

Efficiently Updating ECG-Based Biometric Authentication Based on Incremental Learning

Recently, the interest in biometric authentication based on electrocardiograms (ECGs) has increased. Nevertheless, the ECG signal of a person may vary according to factors such as the emotional or physical state, thus hindering authentication. We propose an adaptive ECG-based authentication method that performs incremental learning to identify ECG signals from a subject under a variety of measurement conditions. An incremental support vector machine (SVM) is adopted for authentication implementing incremental learning. We collected ECG signals from 11 subjects during 10 min over six days and used the data from days 1 to 5 for incremental learning, and those from day 6 for testing. The authentication results show that the proposed system consistently reduces the false acceptance rate from 6.49% to 4.39% and increases the true acceptance rate from 61.32% to 87.61% per single ECG wave after incremental learning using data from the five days. In addition, the authentication results tested using data obtained a day after the latest training show the false acceptance rate being within reliable range (3.5–5.33%) and improvement of the true acceptance rate (70.05–87.61%) over five days.

Data-driven adaptive modeling method for industrial processes and its application in flotation reagent control

In real industrial processes, new process "excitation" patterns that largely deviate from previously collected training data will appear due to disturbances caused by process inputs. To reduce model mismatch, it is important for a data-driven process model to adapt to new process "excitation" patterns. Although efforts have been devoted to developing adaptive process models to deal with this problem, few studies have attempted to develop an adaptive process model that can incrementally learn new process "excitation" patterns without performance degradation on old patterns. In this study, efforts are devoted to enabling data-driven process models with incremental learning ability. First, a novel incremental learning method is proposed for process model updating. Second, an adaptive neural network process model is developed based on the novel incremental learning method. Third, a nonlinear model predictive control based on the adaptive process model is implemented and applied for flotation reagent control. Experiments based on historical data provide evidence that the newly developed adaptive process model can accommodate new process "excitation" patterns and preserve its performance on old patterns. Furthermore, industry experiments carried out in a real-world lead-zinc froth flotation plant provide industrial evidence and show that the newly designed controller is promising for practical flotation reagent control.

An Appraisal of Incremental Learning Methods

As a special case of machine learning, incremental learning can acquire useful knowledge from incoming data continuously while it does not need to access the original data. It is expected to have the ability of memorization and it is regarded as one of the ultimate goals of artificial intelligence technology. However, incremental learning remains a long term challenge. Modern deep neural network models achieve outstanding performance on stationary data distributions with batch training. This restriction leads to catastrophic forgetting for incremental learning scenarios since the distribution of incoming data is unknown and has a highly different probability from the old data. Therefore, a model must be both plastic to acquire new knowledge and stable to consolidate existing knowledge. This review aims to draw a systematic review of the state of the art of incremental learning methods. Published reports are selected from Web of Science, IEEEXplore, and DBLP databases up to May 2020. Each paper is reviewed according to the types: architectural strategy, regularization strategy and rehearsal and pseudo-rehearsal strategy. We compare and discuss different methods. Moreover, the development trend and research focus are given. It is concluded that incremental learning is still a hot research area and will be for a long period. More attention should be paid to the exploration of both biological systems and computational models.