Conformalized quantile regression, a cutting-edge and model-agnostic algorithm, has emerged as a recent innovation to generate valid prediction intervals on finite samples while addressing heteroscedasticity. It starts by employing quantile regression to estimate conditional quantiles. Subsequently, these estimated conditional quantiles undergo a rectification process using conformal prediction. Under the assumption of exchangeability, a slightly weaker form of independent and identically distributed (i.i.d.) data, the resulting prediction intervals are valid in finite samples. However, a drawback of the proposed conformalization step is identified: it lacks the capacity to adapt to heteroscedasticity due to its independence from the input. To overcome this limitation, we propose an improvement that involves partitioning the covariates space into clusters, assigning higher weights to features with greater predictive power. Following that, within each cluster, a conformal step is applied, leveraging a rectification that is reliant on the input cluster-wise. To demonstrate the superiority of our improved version over the classic version of conformalized quantile regression, we conducted a comprehensive comparison of their respective prediction intervals using synthetic data.

This paper introduces a novel model-agnostic algorithm called adaptive ensemble batch multi-input multi-output conformalized quantile regression (AEnbMIMOCQR) that enables forecasters to generate multi-step ahead prediction intervals for a fixed pre-specified miscoverage rate α in a distribution-free manner. Our method is grounded on conformal prediction principles, however, it does not require data splitting and provides close to exact coverage even when the data is not exchangeable. Moreover, the resulting prediction intervals, besides being empirically valid along the forecast horizon, do not neglect heteroscedasticity. AEnbMIMOCQR is designed to be robust to distribution shifts, which means that its prediction intervals remain reliable over an unlimited period of time, without entailing retraining or imposing unrealistic strict assumptions on the data-generating process. Through methodically experimentation, we demonstrate that our approach outperforms other competitive methods on both real-world and synthetic datasets. The code used in the experimental part and a tutorial on how to use AEnbMIMOCQR can be found at the following GitHub repository: https://github.com/Quilograma/AEnbMIMOCQR.

Service robots are expected to be more autonomous and work effectively in human-centric environments. This implies that robots should have special capabilities, such as learning from past experiences and real-time object category recognition. This paper proposes an open-ended 3D object recognition system which concurrently learns both the object categories and the statistical features for encoding objects. In particular, we propose an extension of Latent Dirichlet Allocation to learn structural semantic features (i.e., visual topics), from low-level feature co-occurrences, for each category independently. Moreover, topics in each category are discovered in an unsupervised fashion and are updated incrementally using new object views. In this way, the advantages of both the (hand-crafted) local features and the (learned) structural semantic features have been considered and combined in an efficient way. An extensive set of experiments has been performed to assess the performance of the proposed Local-LDA in terms of descriptiveness, scalability, and computation time. Experimental results show that the overall classification performance obtained with Local-LDA is clearly better than the best performances obtained with the state-of-the-art approaches. Moreover, the best scalability, in terms of number of learned categories, was obtained with the proposed Local-LDA approach, closely followed by a Bag-of-Words (BoW) approach. Concerning computation time, the best result was obtained with BoW, immediately followed by the Local-LDA approach.

Service robots are expected to autonomously and efficiently work in human-centric environments. For this type of robots, object perception and manipulation are challenging tasks due to need for accurate and real-time response. This paper presents an interactive open-ended learning approach to recognize multiple objects and their grasp affordances concurrently. This is an important contribution in the field of service robots since no matter how extensive the training data used for batch learning, a robot might always be confronted with an unknown object when operating in human-centric environments. The paper describes the system architecture and the learning and recognition capabilities. Grasp learning associates grasp configurations (i.e., end-effector positions and orientations) to grasp affordance categories. The grasp affordance category and the grasp configuration are taught through verbal and kinesthetic teaching, respectively. A Bayesian approach is adopted for learning and recognition of object categories and an instance-based approach is used for learning and recognition of affordance categories. An extensive set of experiments has been performed to assess the performance of the proposed approach regarding recognition accuracy, scalability and grasp success rate on challenging datasets and real-world scenarios.

There is growing need for robots that can interact with people in everyday situations. For service robots, it is not reasonable to assume that one can pre-program all object categories. Instead, apart from learning from a batch of labelled training data, robots should continuously update and learn new object categories while working in the environment. This paper proposes a cognitive architecture designed to create a concurrent 3D object category learning and recognition in an interactive and open-ended manner. In particular, this cognitive architecture provides automatic perception capabilities that will allow robots to detect objects in highly crowded scenes and learn new object categories from the set of accumulated experiences in an incremental and open-ended way. Moreover, it supports constructing the full model of an unknown object in an on-line manner and predicting next best view for improving object detection and manipulation performance. We provide extensive experimental results demonstrating system performance in terms of recognition, scalability, next-best-view prediction and real-world robotic applications.

To deploy a robot in a human-centric environment, it is important that the robot is able to continuously acquire and update object categories while working in the environment. Therefore, autonomous robots must have the ability to continuously execute learning and recognition in a concurrent or interleaved fashion. One of the main challenges in unconstrained human environments is to cope with the effects of context change. This paper presents two main contributions: (i) an approach for evaluating open-ended object category learning and recognition methods in multi-context scenarios; (ii) evaluation of different object category learning and recognition approaches regarding their ability to cope with the effects of context change. Off-line evaluation approaches such as cross-validation do not comply with the simultaneous nature of learning and recognition. A teaching protocol, supporting context change, was therefore designed and used in this work for experimental evaluation. Seven learning and recognition approaches were evaluated and compared using the protocol. The best performance, in terms of number of learned categories, was obtained with a recently proposed local variant of Latent Dirichlet Allocation (LDA), closely followed by a Bag-of-Words (BoW) approach. In terms of adaptability, i.e. coping with context change, the best result was obtained with BoW, immediately followed by the local LDA variant.

Object representation is one of the most challenging tasks in robotics because it must provide reliable information in real-time to enable the robot to physically interact with the objects in its environment. To ensure reliability, a global object descriptor must be computed based on a unique and repeatable object reference frame. Moreover, the descriptor should contain enough information enabling to recognize the same or similar objects seen from different perspectives. This paper presents a new object descriptor named Global Orthographic Object Descriptor (GOOD) designed to be robust, descriptive and efficient to compute and use. The performance of the proposed object descriptor is compared with the main state-of-the-art descriptors. Experimental results show that the overall classification performance obtained with GOOD is comparable to the best performances obtained with the state-of-the-art descriptors. Concerning memory and computation time, GOOD clearly outperforms the other descriptors. Therefore, GOOD is especially suited for real-time applications.

Most robots lack the ability to learn new objects from past experiences. To migrate a robot to a new environment one must often completely re-generate the knowledge- base that it is running with. Since in open-ended domains the set of categories to be learned is not predefined, it is not feasible to assume that one can pre-program all object categories required by robots. Therefore, autonomous robots must have the ability to continuously execute learning and recognition in a concurrent and interleaved fashion. This paper proposes an open-ended 3D object recognition system which concurrently learns both the object categories and the statistical features for encoding objects. In particular, we propose an extension of Latent Dirichlet Allocation to learn structural semantic features (i.e. topics) from low-level feature co-occurrences for each category independently. Moreover, topics in each category are discovered in an unsupervised fashion and are updated incrementally using new object views. The approach contains similarities with the organization of the visual cortex and builds a hierarchy of increasingly sophisticated representations. Results show the fulfilling performance of this approach on different types of objects. Moreover, this system demonstrates the capability of learning from few training examples and competes with state-of-the-art systems.

In open-ended domains, robots must continuously learn new object categories. When the training sets are created offline, it is not possible to ensure their representativeness with respect to the object categories and features the system will find when operating online. In the Bag of Words model, visual codebooks are usually constructed from training sets created offline. This might lead to non-discriminative visual words and, as a consequence, to poor recognition performance. This paper proposes a visual object recognition system which concurrently learns in an incremental and online fashion both the visual object category representations as well as the codebook words used to encode them. The codebook is defined using Gaussian Mixture Models which are updated using new object views. The approach contains similarities with the human visual object recognition system: evidence suggests that the development of recognition capabilities occurs on multiple levels and is sustained over large periods of time. Results show that the proposed system with concurrent learning of object categories and codebooks is capable of learning more categories, requiring less examples, and with similar accuracies, when compared to the classical Bag of Words approach using codebooks constructed offline.

This paper addresses the problem of grounding semantic representations in intelligent service robots. In particular, this work contributes to addressing two important aspects, namely the anchoring of object symbols into the perception of the objects and the grounding of object category symbols into the perception of known instances of the categories. The paper discusses memory requirements for storing both semantic and perceptual data and, based on the analysis of these requirements, proposes an approach based on two memory components, namely a semantic memory and a perceptual memory. The perception, memory, learning and interaction capabilities, and the perceptual memory, are the main focus of the paper. Three main design options address the key computational issues involved in processing and storing perception data: a lightweight, NoSQL database, is used to implement the perceptual memory; a thread-based approach with zero copy transport of messages is used in implementing the modules; and a multiplexing scheme, for the processing of the different objects in the scene, enables parallelization. The system is designed to acquire new object categories in an incremental and open-ended way based on user-mediated experiences. The system is fully integrated in a broader robot system comprising low-level control and reactivity to high-level reasoning and learning.

Intelligent service robots should be able to improve their knowledge from accumulated experiences through continuous interaction with the environment, and in particular with humans. A human user may guide the process of experience acquisition, teaching new concepts, or correcting insufficient or erroneous concepts through interaction. This paper reports on work towards interactive learning of objects and robot activities in an incremental and open-ended way. In particular, this paper addresses human-robot interaction and experience gathering. The robot's ontology is extended with concepts for representing human-robot interactions as well as the experiences of the robot. The human-robot interaction ontology includes not only instructor teaching activities but also robot activities to support appropriate feedback from the robot. Two simplified interfaces are implemented for the different types of instructions including the teach instruction, which triggers the robot to extract experiences. These experiences, both in the robot activity domain and in the perceptual domain, are extracted and stored in memory, and they are used as input for learning methods. The functionalities described above are completely integrated in a robot architecture, and are demonstrated in a PR2 robot.

Biomedical signals are in general non-linear and non-stationary which renders them difficult to analyze with classical time series analysis techniques. Empirical Mode Decomposition (EMD) in conjunction with a Hilbert spectral transform, together called Hilbert-Huang Transform, is ideally suited to extract informative components which are characteristic of underlying biological or physiological processes. The method is fully adaptive and generates a complete set of orthogonal basis functions, called Intrinsic Mode Functions (IMFs), in a purely data-driven manner. Amplitude and frequency of IMFs may vary over time which renders them different from conventional basis systems and ideally suited to study non-linear and non-stationary time series. However, biomedical time series are often recorded over long time periods. This generates the need for efficient EMD algorithms which can analyze the data in real time. No such algorithms yet exist which are robust, efficient and easy to implement. The contribution shortly reviews the technique of EMD and related algorithms and develops an on-line variant, called Sliding Empirical Mode Decomposition (SEMD), which is shown to perform well on large scale time series.

This paper is focused on proving the concept that Visually Evoked Potential (VEP) signals registered during experiments with mental task execution can be used for discrimination of individuals. The viability of the VEP-based person identification was successfully tested for a data base of 13 persons. Among various classifiers tested, Support Vector Machine (SVM) with Radial Basis Function (RBF) exhibits the best performance. Our study revealed that the duration of the VEPs required for a reliable identification is crucial for real time implementation of the proposed biometry paradigm. A post processing procedure is formulated in order to overcome the problem of static classification. Our ultimate goal is not to compete with the conventional biometry (such as fingerprint, iris or palm recognition systems) but to design a VEP -based biometry modality as a supplement (“a second opinion”) in clinical conditions. It could be used to verify a patient's identity in medical records, prior to medical procedures or to detect early in advance abnormal mental states of the patient.

Due to external stimuli, biomedical signals are in general non-linear and non-stationary. Intelligent signal processing is crucial to unravel the information content buried in biomedical time series. Empirical Mode Decomposition is ideally suited to extract all pure oscillatory modes which are contained in the signal. These modes, called Intrinsic Mode Functions (IMFs), represent a complete set of locally orthogonal basis functions with time-varying amplitude and frequency. The contribution discusses the application of an online variant, called SEMD, to non-stationary biomedical time series recorded during neuromonitoring.

Due to external stimuli, biomedical signals are in general non-linear and non-stationary. Empirical Mode Decomposition in conjunction with a Hilbert spectral transform, together called Hilbert-Huang Transform, is ideally suited to extract essential components which are characteristic of the underlying biological or physiological processes. The method is fully adaptive and generates the basis to represent the data solely from these data and based on them. The basis functions, called Intrinsic Mode Functions (IMFs) represent a complete set of locally orthogonal basis functions whose amplitude and frequency may vary over time. The contribution reviews the technique of EMD and related algorithms and discusses illustrative applications.

MOTIVATION: Modern machine learning methods based on matrix decomposition techniques, like independent component analysis (ICA) or non-negative matrix factorization (NMF), provide new and efficient analysis tools which are currently explored to analyze gene expression profiles. These exploratory feature extraction techniques yield expression modes (ICA) or metagenes (NMF). These extracted features are considered indicative of underlying regulatory processes. They can as well be applied to the classification of gene expression datasets by grouping samples into different categories for diagnostic purposes or group genes into functional categories for further investigation of related metabolic pathways and regulatory networks. RESULTS: In this study we focus on unsupervised matrix factorization techniques and apply ICA and sparse NMF to microarray datasets. The latter monitor the gene expression levels of human peripheral blood cells during differentiation from monocytes to macrophages. We show that these tools are able to identify relevant signatures in the deduced component matrices and extract informative sets of marker genes from these gene expression profiles. The methods rely on the joint discriminative power of a set of marker genes rather than on single marker genes. With these sets of marker genes, corroborated by leave-one-out or random forest cross-validation, the datasets could easily be classified into related diagnostic categories. The latter correspond to either monocytes versus macrophages or healthy vs Niemann Pick C disease patients.

Nonnegative Matrix Factorization (NMF) has proven to be a useful tool for the analysis of nonnegative multivariate data. Gene expression profiles naturally conform to assumptions about data formats raised by NMF. However, it is known not to lead to unique results concerning the component signals extracted. In this paper we consider an extension of the NMF algorithm which provides unique solutions whenever the underlying component signals are sufficiently sparse. A new sparseness measure is proposed most appropriate to suitably transformed gene expression profiles. The resulting fitness function is discontinuous and exhibits many local minima, hence we use a genetic algorithm for its optimization. The algorithm is applied to toy data to investigate its properties as well as to a microarray data set related to Pseudo-Xanthoma Elasticum (PXE).

In this work an automatic assignment tool for estimated independent components within an independent component analysis is presented. The tool is applied to the problem of removing the water resonance and related artifacts from multi-dimensional proton NMR spectra. The algorithm uses local PCA to approximate the water artifact and defines a suitable cost function which is optimized using simulated annealing. The blind extraction of artifact-related source signals is effected by a recently developed algorithm called dAMUSE.

We present a method based on singular spectrum analysis to remove ocular artifacts (EOG) from an electroencephalogram (EEC). After embedding the EEG signals in a feature space of time-delayed coordinates, feature vectors are clustered and the principal components (PCs) are computed locally within each cluster. Then we assume that the EOG artifact is associated with the PCs belonging to the largest eigenvalues. We incorporate a minimum description length (IMDL) criterion to estimate the number of eigenvectors needed to represent the EOG artifact faithfully. The EOG signal thus extracted is then subtracted from the original EEG signal to obtain the corrected EEG signal we are interested in.

We present a denoising algorithm for enhancing noisy signals based on local independent component analysis (ICA). This is done by applying ICA to the signal in localized delayed coordinates. The components resembling the signals can be detected by various criteria depending on the nature of the signal. Estimators of kurtosis or the variance of the autocorrelation have been considered. The algorithm proposed can favorably be applied to the problem of denoising multidimensional data like images or fMRI data sets. In comparison to denoising algorithms using wavelets, Wiener filters and kernel PCA the local PCA and ICA algorithms perform considerably better. We provide applications of the algorithm to images and the analysis of protein NMR spectra.

Multidimensional /sup 1/H NMR spectra of biomolecules dissolved in light water are contaminated by an intense water artifact. Generalized eigenvalue decomposition methods using congruent matrix pencils are used to separate the water artefact from the protein spectra. Due to the statistical separation process, however, noise is introduced into the reconstructed spectra. Hence Kernel-based denoising techniques are discussed to obtain noise- and artifact-free 2D NOESY NMR spectra of proteins.

In this work a modified version of AMUSE, called dAMUSE, is proposed. The main modification consists in increasing the dimension of the data vectors by joining delayed versions of the observed mixed signals. With the new data a matrix pencil is computed and its generalized eigendecomposition is performed as in AMUSE. We will show that in this case the output (or independent) signals are filtered versions of the source signals. Some numerical simulations using artificially mixed signals as well as biological data (RR and QT intervals of Electrocardiogram) are presented.

The blind source separation might be addressed as a generalized eigenvalue problem. A complete formulation using linear algebra statements shows that the required matrix pencil might be computed at the input and at the output of a simple linear filter. This method is evaluated against FastICA algorithm.

Typically, perceptual audio coders have followed a subband or transform coding scheme with forward-adaptive quantization. In this letter we present an alternative scheme which uses backward-adaptive quantization, we discuss the effects of this strategy on perceptual coding and show that it can be successfully applied.

The blind source separation problem was recently addressed as a generalized eigenvalue decompisition (GSVD) problem. This work is concerned with the formalization of the method using linear algebra statements. This mathematical framework also reveals a new method to compute the two correlation matrices used in the approach.

A set of parameters to assess the objective quality of visualizations of a voxel-based data set, produced using a ray caster, is proposed as a first step toward the evaluation of the overall quality of these visualizations. Results obtained using synthetic data and a simple implementation of a ray caster are presented. The final goal of this evaluation is the computation of "confidence indices" that could offer the user a "guided visualization", i.e. allow him/her to decide what are the "best" visualizations of a data set.