Trustworthy ML systems should not only return accurate predictions, but also a reliable representation of their uncertainty. Bayesian methods are commonly used to quantify both aleatoric and epistemic uncertainty, but alternative approaches, such as evidential deep learning methods, have become popular in recent years. The latter group of methods in essence extends empirical risk minimization (ERM) for predicting second-order probability distributions over outcomes, from which measures of epistemic (and aleatoric) uncertainty can be extracted. This paper presents novel theoretical insights of evidential deep learning, highlighting the difficulties in optimizing second-order loss functions and interpreting the resulting epistemic uncertainty measures. With a systematic setup that covers a wide range of approaches for classification, regression and counts, it provides novel insights into issues of identifiability and convergence in second-order loss minimization, and the relative (rather than absolute) nature of epistemic uncertainty measures.

Continual learning is a subfield of machine learning, which aims to allow machine learning models to continuously learn on new data, by accumulating knowledge without forgetting what was learned in the past. In this work, we take a step back, and ask: "Why should one care about continual learning in the first place?". We set the stage by examining recent continual learning papers published at four major machine learning conferences, and show that memory-constrained settings dominate the field. Then, we discuss five open problems in machine learning, and even though they might seem unrelated to continual learning at first sight, we show that continual learning will inevitably be part of their solution. These problems are model editing, personalization and specialization, on-device learning, faster (re-)training and reinforcement learning. Finally, by comparing the desiderata from these unsolved problems and the current assumptions in continual learning, we highlight and discuss four future directions for continual learning research. We hope that this work offers an interesting perspective on the future of continual learning, while displaying its potential value and the paths we have to pursue in order to make it successful. This work is the result of the many discussions the authors had at the Dagstuhl seminar on Deep Continual Learning, in March 2023.

Facing the dynamic complex cyber environments, internal and external cyber threat intelligence, and the increasing risk of cyber-attack, knowledge graphs show great application potential in the cyber security area because of their capabilities in knowledge aggregation, representation, management, and reasoning. However, while most research has focused on how to develop a complete knowledge graph, it remains unclear how to apply the knowledge graph to solve industrial real challenges in cyber-attack and defense scenarios. In this review, we provide a brief overview of the basic concepts, schema, and construction approaches for the cyber security knowledge graph. To facilitate future research on cyber security knowledge graphs, we also present a curated collection of datasets and open-source libraries on the knowledge construction and information extraction task. In the major part of this article, we conduct a comparative review of the different works that elaborate on the recent progress in the application scenarios of the cyber security knowledge graph. Furthermore, a novel comprehensive classification framework is created to describe the connected works from nine primary categories and eighteen subcategories. Finally, we have a thorough outlook on several promising research directions based on the discussion of existing research flaws.

It is well known that accurate probabilistic predictors can be trained through empirical risk minimisation with proper scoring rules as loss functions. While such learners capture so-called aleatoric uncertainty of predictions, various machine learning methods have recently been developed with the goal to let the learner also represent its epistemic uncertainty, i.e., the uncertainty caused by a lack of knowledge and data. An emerging branch of the literature proposes the use of a second-order learner that provides predictions in terms of distributions on probability distributions. However, recent work has revealed serious theoretical shortcomings for second-order predictors based on loss minimisation. In this paper, we generalise these findings and prove a more fundamental result: There seems to be no loss function that provides an incentive for a second-order learner to faithfully represent its epistemic uncertainty in the same manner as proper scoring rules do for standard (first-order) learners. As a main mathematical tool to prove this result, we introduce the generalised notion of second-order scoring rules.

This paper presents our findings from participating in the SMM4H Shared Task 2021. We addressed Named Entity Recognition (NER) and Text Classification. To address NER we explored BiLSTM-CRF with Stacked Heterogeneous Embeddings and linguistic features. We investigated various machine learning algorithms (logistic regression, Support Vector Machine (SVM) and Neural Networks) to address text classification. Our proposed approaches can be generalized to different languages and we have shown its effectiveness for English and Spanish. Our text classification submissions (team:MIC-NLP) have achieved competitive performance with F1-score of $0.46$ and $0.90$ on ADE Classification (Task 1a) and Profession Classification (Task 7a) respectively. In the case of NER, our submissions scored F1-score of $0.50$ and $0.82$ on ADE Span Detection (Task 1b) and Profession Span detection (Task 7b) respectively.

In this work, we combine the two paradigms: Federated Learning (FL) and Continual Learning (CL) for text classification task in cloud-edge continuum. The objective of Federated Continual Learning (FCL) is to improve deep learning models over life time at each client by (relevant and efficient) knowledge transfer without sharing data. Here, we address challenges in minimizing inter-client interference while knowledge sharing due to heterogeneous tasks across clients in FCL setup. In doing so, we propose a novel framework, Federated Selective Inter-client Transfer (FedSeIT) which selectively combines model parameters of foreign clients. To further maximize knowledge transfer, we assess domain overlap and select informative tasks from the sequence of historical tasks at each foreign client while preserving privacy. Evaluating against the baselines, we show improved performance, a gain of (average) 12.4\% in text classification over a sequence of tasks using five datasets from diverse domains. To the best of our knowledge, this is the first work that applies FCL to NLP.

Robotic systems operating at the edge require efficient online learning algorithms that can continuously adapt to changing environments while processing streaming sensory data. Traditional backpropagation, while effective, conflicts with biological plausibility principles and may be suboptimal for continuous adaptation scenarios. The Predictive Coding (PC) framework offers a biologically plausible alternative with local, Hebbian-like update rules, making it suitable for neuromorphic hardware implementation. However, PC's main limitation is its computational overhead due to multiple inference iterations during training. We present Predictive Coding Network with Temporal Amortization (PCN-TA), which preserves latent states across temporal frames. By leveraging temporal correlations, PCN-TA significantly reduces computational demands while maintaining learning performance. Our experiments on the COIL-20 robotic perception dataset demonstrate that PCN-TA achieves 10% fewer weight updates compared to backpropagation and requires 50% fewer inference steps than baseline PC networks. These efficiency gains directly translate to reduced computational overhead for moving another step toward edge deployment and real-time adaptation support in resource-constrained robotic systems. The biologically-inspired nature of our approach also makes it a promising candidate for future neuromorphic hardware implementations, enabling efficient online learning at the edge.

Multifunctional mesoporous silica nanoparticles (MSN) have attracted substantial attention with regard to their high potential for targeted drug delivery. For future clinical applications it is crucial to address safety concerns and understand the potential immunotoxicity of these nanoparticles. In this study, we assess the biocompatibility and functionality of multifunctional MSN in freshly isolated, primary murine immune cells. We show that the functionalized silica nanoparticles are rapidly and efficiently taken up into the endosomal compartment by specialized antigen-presenting cells such as dendritic cells. The silica nanoparticles showed a favorable toxicity profile and did not affect the viability of primary immune cells from the spleen in relevant concentrations. Cargo-free MSN induced only very low immune responses in primary cells as determined by surface expression of activation markers and release of pro-inflammatory cytokines such as Interleukin-6, -12 and -1\beta. In contrast, when surface-functionalized MSN with a pH-responsive polymer capping were loaded with an immune-activating drug, the synthetic Toll-like receptor 7 agonist R848, a strong immune response was provoked. We thus demonstrate that MSN represent an efficient drug delivery vehicle to primary immune cells that is both non-toxic and non-inflammagenic, which is a prerequisite for the use of these particles in biomedical applications.

Though word embeddings and topics are complementary representations, several past works have only used pretrained word embeddings in (neural) topic modeling to address data sparsity in short-text or small collection of documents. This work presents a novel neural topic modeling framework using multi-view embedding spaces: (1) pretrained topic-embeddings, and (2) pretrained word-embeddings (context insensitive from Glove and context-sensitive from BERT models) jointly from one or many sources to improve topic quality and better deal with polysemy. In doing so, we first build respective pools of pretrained topic (i.e., TopicPool) and word embeddings (i.e., WordPool). We then identify one or more relevant source domain(s) and transfer knowledge to guide meaningful learning in the sparse target domain. Within neural topic modeling, we quantify the quality of topics and document representations via generalization (perplexity), interpretability (topic coherence) and information retrieval (IR) using short-text, long-text, small and large document collections from news and medical domains. Introducing the multi-source multi-view embedding spaces, we have shown state-of-the-art neural topic modeling using 6 source (high-resource) and 5 target (low-resource) corpora.

The problem of selecting an algorithm that appears most suitable for a specific instance of an algorithmic problem class, such as the Boolean satisfiability problem, is called instance-specific algorithm selection. Over the past decade, the problem has received considerable attention, resulting in a number of different methods for algorithm selection. Although most of these methods are based on machine learning, surprisingly little work has been done on meta learning, that is, on taking advantage of the complementarity of existing algorithm selection methods in order to combine them into a single superior algorithm selector. In this paper, we introduce the problem of meta algorithm selection, which essentially asks for the best way to combine a given set of algorithm selectors. We present a general methodological framework for meta algorithm selection as well as several concrete learning methods as instantiations of this framework, essentially combining ideas of meta learning and ensemble learning. In an extensive experimental evaluation, we demonstrate that ensembles of algorithm selectors can significantly outperform single algorithm selectors and have the potential to form the new state of the art in algorithm selection.

The article is concerned with the problem of multi-step financial time series forecasting of Foreign Exchange (FX) rates. To address this problem, we introduce a regression network termed RegPred Net. The exchange rate to forecast is treated as a stochastic process. It is assumed to follow a generalization of Brownian motion and the mean-reverting process referred to as the generalized Ornstein-Uhlenbeck (OU) process, with time-dependent coefficients. Using past observed values of the input time series, these coefficients can be regressed online by the cells of the first half of the network (Reg). The regressed coefficients depend only on - but are very sensitive to - a small number of hyperparameters required to be set by a global optimization procedure for which, Bayesian optimization is an adequate heuristic. Thanks to its multi-layered architecture, the second half of the regression network (Pred) can project time-dependent values for the OU process coefficients and generate realistic trajectories of the time series. Predictions can be easily derived in the form of expected values estimated by averaging values obtained by Monte Carlo simulation. The forecasting accuracy on a 100 days horizon is evaluated for several of the most important FX rates such as EUR/USD, EUR/CNY, and EUR/GBP. Our experimental results show that the RegPred Net significantly outperforms ARMA, ARIMA, LSTMs, and Autoencoder-LSTM models in terms of metrics measuring the absolute error (RMSE) and correlation between predicted and actual values (Pearson R, R-squared, MDA). Compared to black-box deep learning models such as LSTM, RegPred Net has better interpretability, simpler structure, and fewer parameters.

In recent years, Explainable AI (xAI) attracted a lot of attention as various countries turned explanations into a legal right. xAI allows for improving models beyond the accuracy metric by, e.g., debugging the learned pattern and demystifying the AI's behavior. The widespread use of xAI brought new challenges. On the one hand, the number of published xAI algorithms underwent a boom, and it became difficult for practitioners to select the right tool. On the other hand, some experiments did highlight how easy data scientists could misuse xAI algorithms and misinterpret their results. To tackle the issue of comparing and correctly using feature importance xAI algorithms, we propose Compare-xAI, a benchmark that unifies all exclusive functional testing methods applied to xAI algorithms. We propose a selection protocol to shortlist non-redundant functional tests from the literature, i.e., each targeting a specific end-user requirement in explaining a model. The benchmark encapsulates the complexity of evaluating xAI methods into a hierarchical scoring of three levels, namely, targeting three end-user groups: researchers, practitioners, and laymen in xAI. The most detailed level provides one score per test. The second level regroups tests into five categories (fidelity, fragility, stability, simplicity, and stress tests). The last level is the aggregated comprehensibility score, which encapsulates the ease of correctly interpreting the algorithm's output in one easy to compare value. Compare-xAI's interactive user interface helps mitigate errors in interpreting xAI results by quickly listing the recommended xAI solutions for each ML task and their current limitations. The benchmark is made available at https://karim-53.github.io/cxai/

A recent study by Panchagnula et al. [J. Chem. Phys. 161, 054308 (2024)] illustrated the non-concordance of a variety of electronic structure methods at describing the symmetric double-well potential expected along the anisotropic direction of the endofullerene Ne@C$_{70}$. In this article we delve deeper into the difficulties of accurately capturing the dispersion interaction within this system, scrutinising a variety of state-of-the-art density-functional approximations (DFAs) and dispersion corrections (DCs). We identify rigorous criteria for the double-well potential and compare the shapes, barrier heights, and minima positions obtained with the DFAs and DCs to the correlated wavefunction data in the previous study, alongside new coupled-cluster calculations. We show that many of the DFAs are extremely sensitive to the numerical integration grid used, and note that the choice of DC is not independent of the DFA. Functionals with many empirical parameters tuned for main-group thermochemistry do not necessarily result in a reasonable PES, while improved performance can be obtained using nearly dispersionless DFAs with very few empirical parameters and allowing the DC to compensate. We pose the Ne@C$_{70}$ system as a challenge to functional developers and as a diagnostic system for testing dispersion corrections, and reiterate the need for more experimental data for comparison.

Recent years, weather forecasting has gained significant attention. However, accurately predicting weather remains a challenge due to the rapid variability of meteorological data and potential teleconnections. Current spatiotemporal forecasting models primarily rely on convolution operations or sliding windows for feature extraction. These methods are limited by the size of the convolutional kernel or sliding window, making it difficult to capture and identify potential teleconnection features in meteorological data. Additionally, weather data often involve non-rigid bodies, whose motion processes are accompanied by unpredictable deformations, further complicating the forecasting task. In this paper, we propose the GMG model to address these two core challenges. The Global Focus Module, a key component of our model, enhances the global receptive field, while the Motion Guided Module adapts to the growth or dissipation processes of non-rigid bodies. Through extensive evaluations, our method demonstrates competitive performance across various complex tasks, providing a novel approach to improving the predictive accuracy of complex spatiotemporal data.

Active and reliable electrocatalysts are fundamental to renewable energy technologies. PdCoO2 has recently been recognized as a promising catalyst template for the hydrogen evolution reaction (HER) in acidic media thanks to the formation of active PdHx. In this article, we monitor the transformation of single PdCoO2 particles during HER, and confirm their almost complete transformation to PdHx with sub-millimeter depths and cracks throughout the particles. Using operando mass spectrometry, Co dissolution is observed under reductive potentials, leading to PdHx formation, whereas the dissolution partial current is found to be 0.1 % of the HER current. The formation of PdHx is confirmed through secondary ion mass spectrometry and quantitatively analyzed by atom probe tomography, enabled by isotope labelling of hydrogen using heavy water. Despite dry storage and high vacuum during sample preparations, an overall composition of PdD0.28 is measured for the PdHx sample, with separation between alpha- (D-poor) and beta- (D-rich) PdHx phases. The PdHx phase formed on PdCoO2 particles is stable for a wide electrochemical potential window, until Pd dissolution is observed at open circuit potentials. Our findings highlight the critical role of a templated growth method in obtaining stabilized PdHx, enabling efficient HER without the commonly slow activation processes observed in Pd. This offers insights into the design of more efficient electrocatalysts for renewable energy technologies.

We address two challenges in topic models: (1) Context information around words helps in determining their actual meaning, e.g., "networks" used in the contexts "artificial neural networks" vs. "biological neuron networks". Generative topic models infer topic-word distributions, taking no or only little context into account. Here, we extend a neural autoregressive topic model to exploit the full context information around words in a document in a language modeling fashion. The proposed model is named as iDocNADE. (2) Due to the small number of word occurrences (i.e., lack of context) in short text and data sparsity in a corpus of few documents, the application of topic models is challenging on such texts. Therefore, we propose a simple and efficient way of incorporating external knowledge into neural autoregressive topic models: we use embeddings as a distributional prior. The proposed variants are named as DocNADEe and iDocNADEe. We present novel neural autoregressive topic model variants that consistently outperform state-of-the-art generative topic models in terms of generalization, interpretability (topic coherence) and applicability (retrieval and classification) over 7 long-text and 8 short-text datasets from diverse domains.

This paper describes our system (MIC-CIS) details and results of participation in the fine-grained propaganda detection shared task 2019. To address the tasks of sentence (SLC) and fragment level (FLC) propaganda detection, we explore different neural architectures (e.g., CNN, LSTM-CRF and BERT) and extract linguistic (e.g., part-of-speech, named entity, readability, sentiment, emotion, etc.), layout and topical features. Specifically, we have designed multi-granularity and multi-tasking neural architectures to jointly perform both the sentence and fragment level propaganda detection. Additionally, we investigate different ensemble schemes such as majority-voting, relax-voting, etc. to boost overall system performance. Compared to the other participating systems, our submissions are ranked 3rd and 4th in FLC and SLC tasks, respectively.

Lifelong learning has recently attracted attention in building machine learning systems that continually accumulate and transfer knowledge to help future learning. Unsupervised topic modeling has been popularly used to discover topics from document collections. However, the application of topic modeling is challenging due to data sparsity, e.g., in a small collection of (short) documents and thus, generate incoherent topics and sub-optimal document representations. To address the problem, we propose a lifelong learning framework for neural topic modeling that can continuously process streams of document collections, accumulate topics and guide future topic modeling tasks by knowledge transfer from several sources to better deal with the sparse data. In the lifelong process, we particularly investigate jointly: (1) sharing generative homologies (latent topics) over lifetime to transfer prior knowledge, and (2) minimizing catastrophic forgetting to retain the past learning via novel selective data augmentation, co-training and topic regularization approaches. Given a stream of document collections, we apply the proposed Lifelong Neural Topic Modeling (LNTM) framework in modeling three sparse document collections as future tasks and demonstrate improved performance quantified by perplexity, topic coherence and information retrieval task.

Effective and controlled drug delivery systems with on-demand release and targeting abilities have received enormous attention for biomedical applications. Here, we describe a novel enzyme-based cap system for mesoporous silica nanoparticles (MSNs) that is directly combined with a targeting ligand via bio-orthogonal click chemistry. The capping system is based on the pH-responsive binding of an aryl-sulfonamide-functionalized MSN and the enzyme carbonic anhydrase (CA). An unnatural amino acid (UAA) containing a norbornene moiety was genetically incorporated into CA. This UAA allowed for the site-specific bio-orthogonal attachment of even very sensitive targeting ligands such as folic acid and anandamide. This leads to specific receptor-mediated cell and stem cell uptake. We demonstrate the successful delivery and release of the chemotherapeutic agent Actinomycin D to KB cells. This novel nanocarrier concept provides a promising platform for the development of precisely controllable and highly modular theranostic systems.