arXiv Computer Science @arxiv_cs@qoto.org

Data quality is the key factor for the development of trustworthy AI in healthcare. A large volume of curated datasets with controlled confounding factors can help improve the accuracy, robustness and privacy of downstream AI algorithms. However, access to good quality datasets is limited by the technical difficulty of data acquisition and large-scale sharing of healthcare data is hindered by strict ethical restrictions. Data synthesis algorithms, which generate data with a similar distribution as real clinical data, can serve as a potential solution to address the scarcity of good quality data during the development of trustworthy AI. However, state-of-the-art data synthesis algorithms, especially deep learning algorithms, focus more on imaging data while neglecting the synthesis of non-imaging healthcare data, including clinical measurements, medical signals and waveforms, and electronic healthcare records (EHRs). Thus, in this paper, we will review the synthesis algorithms, particularly for non-imaging medical data, with the aim of providing trustworthy AI in this domain. This tutorial-styled review paper will provide comprehensive descriptions of non-imaging medical data synthesis on aspects including algorithms, evaluations, limitations and future research directions.

Recently, distributed semi-supervised learning (DSSL) algorithms have shown their effectiveness in leveraging unlabeled samples over interconnected networks, where agents cannot share their original data with each other and can only communicate non-sensitive information with their neighbors. However, existing DSSL algorithms cannot cope with data uncertainties and may suffer from high computation and communication overhead problems. To handle these issues, we propose a distributed semi-supervised fuzzy regression (DSFR) model with fuzzy if-then rules and interpolation consistency regularization (ICR). The ICR, which was proposed recently for semi-supervised problem, can force decision boundaries to pass through sparse data areas, thus increasing model robustness. However, its application in distributed scenarios has not been considered yet. In this work, we proposed a distributed Fuzzy C-means (DFCM) method and a distributed interpolation consistency regularization (DICR) built on the well-known alternating direction method of multipliers to respectively locate parameters in antecedent and consequent components of DSFR. Notably, the DSFR model converges very fast since it does not involve back-propagation procedure and is scalable to large-scale datasets benefiting from the utilization of DFCM and DICR. Experiments results on both artificial and real-world datasets show that the proposed DSFR model can achieve much better performance than the state-of-the-art DSSL algorithm in terms of both loss value and computational cost.

Neural image compression (NIC) has outperformed traditional image codecs in rate-distortion (R-D) performance. However, it usually requires a dedicated encoder-decoder pair for each point on R-D curve, which greatly hinders its practical deployment. While some recent works have enabled bitrate control via conditional coding, they impose strong prior during training and provide limited flexibility. In this paper we propose Code Editing, a highly flexible coding method for NIC based on semi-amortized inference and adaptive quantization. Our work is a new paradigm for variable bitrate NIC. Furthermore, experimental results show that our method surpasses existing variable-rate methods, and achieves ROI coding and multi-distortion trade-off with a single decoder.

The synergistic drug combinations provide huge potentials to enhance therapeutic efficacy and to reduce adverse reactions. However, effective and synergistic drug combination prediction remains an open question because of the unknown causal disease signaling pathways. Though various deep learning (AI) models have been proposed to quantitatively predict the synergism of drug combinations. The major limitation of existing deep learning methods is that they are inherently not interpretable, which makes the conclusion of AI models un-transparent to human experts, henceforth limiting the robustness of the model conclusion and the implementation ability of these models in the real-world human-AI healthcare. In this paper, we develop an interpretable graph neural network (GNN) that reveals the underlying essential therapeutic targets and mechanism of the synergy (MoS) by mining the sub-molecular network of great importance. The key point of the interpretable GNN prediction model is a novel graph pooling layer, Self-Attention based Node and Edge pool (henceforth SANEpool), that can compute the attention score (importance) of nodes and edges based on the node features and the graph topology. As such, the proposed GNN model provides a systematic way to predict and interpret the drug combination synergism based on the detected crucial sub-molecular network. We evaluate SANEpool on molecular networks formulated by genes from 46 core cancer signaling pathways and drug combinations from NCI ALMANAC drug combination screening data. The experimental results indicate that 1) SANEpool can achieve the current state-of-art performance among other popular graph neural networks; and 2) the sub-molecular network detected by SANEpool are self-explainable and salient for identifying synergistic drug combinations.

In the last few years, several initiatives have been starting to offer access to research outputs data and metadata in an open fashion. The platforms developed by those initiatives are opening up scientific production to the wider public and they can be an invaluable asset for evidence-based policy-making in Science, Technology and Innovation (STI). These resources can indeed facilitate knowledge discovery and help identify available R&D assets and relevant actors within specific research niches of interest. Ideally, to gain a comprehensive view of entire STI ecosystems, the information provided by each of these resources should be combined and analysed accordingly. To ensure so, at least a certain degree of interoperability should be guaranteed across data sources, so that data could be better aggregated and complemented and that evidence provided towards policy-making is more complete and reliable. Here, we study whether this is the case for the case of mapping Climate Action research in the whole Denmark STI ecosystem, by using 4 popular open access STI data sources, namely OpenAire, Open Alex, CORDIS and Kohesio.

Removal or cancellation of noise has wide-spread applications for imaging and acoustics. In every-day-life applications, denoising may even include generative aspects which are unfaithful to the ground truth. For scientific applications, however, denoising must reproduce the ground truth accurately. Here, we show how data can be denoised via a deep convolutional neural network such that weak signals appear with quantitative accuracy. In particular, we study X-ray diffraction on crystalline materials. We demonstrate that weak signals stemming from charge ordering, insignificant in the noisy data, become visible and accurate in the denoised data. This success is enabled by supervised training of a deep neural network with pairs of measured low- and high-noise data. This way, the neural network learns about the statistical properties of the noise. We demonstrate that using artificial noise (such as Poisson and Gaussian) does not yield such quantitatively accurate results. Our approach thus illustrates a practical strategy for noise filtering that can be applied to challenging acquisition problems.

Analog neuromorphic hardware promises fast brain emulation on the one hand and an efficient implementation of novel, brain-inspired computing paradigms on the other. Bridging this spectrum requires flexibly configurable circuits with reliable and reproducible dynamics fostered by an accurate implementation of the targeted neuron and synapse models. This manuscript presents the analog neuron circuits of the mixed-signal accelerated neuromorphic system BrainScaleS-2. They are capable of flexibly and accurately emulating the adaptive exponential leaky integrate-and-fire model equations in combination with both current- and conductance-based synapses, as demonstrated by precisely replicating a wide range of complex neuronal dynamics and firing patterns.

We implement and interpret various supervised learning experiments involving real quadratic fields with class numbers 1, 2 and 3. We quantify the relative difficulties in separating class numbers of matching/different parity from a data-scientific perspective, apply the methodology of feature analysis and principal component analysis, and use symbolic classification to develop machine-learned formulas for class numbers 1, 2 and 3 that apply to our dataset.

Centralized approaches for multi-robot coverage planning problems suffer from the lack of scalability. Learning-based distributed algorithms provide a scalable avenue in addition to bringing data-oriented feature generation capabilities to the table, allowing integration with other learning-based approaches. To this end, we present a learning-based, differentiable distributed coverage planner (D2COPL A N) which scales efficiently in runtime and number of agents compared to the expert algorithm, and performs on par with the classical distributed algorithm. In addition, we show that D2COPlan can be seamlessly combined with other learning methods to learn end-to-end, resulting in a better solution than the individually trained modules, opening doors to further research for tasks that remain elusive with classical methods.

This paper develops a methodology for collaborative human-robot exploration that leverages implicit coordination. Most autonomous single- and multi-robot exploration systems require a remote operator to provide explicit guidance to the robotic team. Few works consider how to embed the human partner alongside robots to provide guidance in the field. A remaining challenge for collaborative human-robot exploration is efficient communication of goals from the human to the robot. In this paper we develop a methodology that implicitly communicates a region of interest from a helmet-mounted depth camera on the human's head to the robot and an information gain-based exploration objective that biases motion planning within the viewpoint provided by the human. The result is an aerial system that safely accesses regions of interest that may not be immediately viewable or reachable by the human. The approach is evaluated in simulation and with hardware experiments in a motion capture arena. Videos of the simulation and hardware experiments are available at: https://youtu.be/7jgkBpVFIoE.

Color image denoising is frequently encountered in various image processing and computer vision tasks. One traditional strategy is to convert the RGB image to a less correlated color space and denoise each channel of the new space separately. However, such a strategy can not fully exploit the correlated information between channels and is inadequate to obtain satisfactory results. To address this issue, this paper proposes a new multi-channel optimization model for color image denoising under the nuclear norm minus Frobenius norm minimization framework. Specifically, based on the block-matching, the color image is decomposed into overlapping RGB patches. For each patch, we stack its similar neighbors to form the corresponding patch matrix. The proposed model is performed on the patch matrix to recover its noise-free version. During the recovery process, a) a weight matrix is introduced to fully utilize the noise difference between channels; b) the singular values are shrunk adaptively without additionally assigning weights. With them, the proposed model can achieve promising results while keeping simplicity. To solve the proposed model, an accurate and effective algorithm is built based on the alternating direction method of multipliers framework. The solution of each updating step can be analytically expressed in closed-from. Rigorous theoretical analysis proves the solution sequences generated by the proposed algorithm converge to their respective stationary points. Experimental results on both synthetic and real noise datasets demonstrate the proposed model outperforms state-of-the-art models.

Ventricular tachycardia (VT) can be one cause of sudden cardiac death affecting 4.25 million persons per year worldwide. A curative treatment is catheter ablation in order to inactivate the abnormally triggering regions. To facilitate and expedite the localization during the ablation procedure, we present two novel localization techniques based on convolutional neural networks (CNNs). In contrast to existing methods, e.g. using ECG imaging, our approaches were designed to be independent of the patient-specific geometries and directly applicable to surface ECG signals, while also delivering a binary transmural position. One method outputs ranked alternative solutions. Results can be visualized either on a generic or patient geometry. The CNNs were trained on a data set containing only simulated data and evaluated both on simulated and clinical test data. On simulated data, the median test error was below 3mm. The median localization error on the clinical data was as low as 32mm. The transmural position was correctly detected in up to 82% of all clinical cases. Using the ranked alternative solutions, the top-3 median error dropped to 20mm on clinical data. These results demonstrate a proof of principle to utilize CNNs to localize the activation source without the intrinsic need of patient-specific geometrical information. Furthermore, delivering multiple solutions can help the physician to find the real activation source amongst more than one possible locations. With further optimization, these methods have a high potential to speed up clinical interventions. Consequently they could decrease procedural risk and improve VT patients' outcomes.

Optical coherence tomography (OCT) imaging from different camera devices causes challenging domain shifts and can cause a severe drop in accuracy for machine learning models. In this work, we introduce a minimal noise adaptation method based on a singular value decomposition (SVDNA) to overcome the domain gap between target domains from three different device manufacturers in retinal OCT imaging. Our method utilizes the difference in noise structure to successfully bridge the domain gap between different OCT devices and transfer the style from unlabeled target domain images to source images for which manual annotations are available. We demonstrate how this method, despite its simplicity, compares or even outperforms state-of-the-art unsupervised domain adaptation methods for semantic segmentation on a public OCT dataset. SVDNA can be integrated with just a few lines of code into the augmentation pipeline of any network which is in contrast to many state-of-the-art domain adaptation methods which often need to change the underlying model architecture or train a separate style transfer model. The full code implementation for SVDNA is available at https://github.com/ValentinKoch/SVDNA.

In modern highly interconnected power grids, automatic generation control (AGC) is crucial in maintaining the stability of the power grid. The dependence of the AGC system on the information and communications technology (ICT) system makes it vulnerable to various types of cyber-attacks. Thus, information flow (IF) analysis and anomaly detection became paramount for preventing cyber attackers from driving the cyber-physical power system (CPPS) to instability. In this paper, the ICT network traffic rules in CPPSs are explored and the frequency domain features of the ICT network traffic are extracted, basically for developing a robust learning algorithm that can learn the normal traffic pattern based on the ResNeSt convolutional neural network (CNN). Furthermore, to overcome the problem of insufficient abnormal traffic labeled samples, transfer learning approach is used. In the proposed data-driven-based method the deep learning model is trained by traffic frequency features, which makes our model robust against AGC's parameters uncertainties and modeling nonlinearities.

Ideological divisions in the United States have become increasingly prominent in daily communication. Accordingly, there has been much research on political polarization, including many recent efforts that take a computational perspective. By detecting political biases in a corpus of text, one can attempt to describe and discern the polarity of that text. Intuitively, the named entities (i.e., the nouns and phrases that act as nouns) and hashtags in text often carry information about political views. For example, people who use the term "pro-choice" are likely to be liberal, whereas people who use the term "pro-life" are likely to be conservative. In this paper, we seek to reveal political polarities in social-media text data and to quantify these polarities by explicitly assigning a polarity score to entities and hashtags. Although this idea is straightforward, it is difficult to perform such inference in a trustworthy quantitative way. Key challenges include the small number of known labels, the continuous spectrum of political views, and the preservation of both a polarity score and a polarity-neutral semantic meaning in an embedding vector of words. To attempt to overcome these challenges, we propose the Polarity-aware Embedding Multi-task learning (PEM) model. This model consists of (1) a self-supervised context-preservation task, (2) an attention-based tweet-level polarity-inference task, and (3) an adversarial learning task that promotes independence between an embedding's polarity dimension and its semantic dimensions. Our experimental results demonstrate that our PEM model can successfully learn polarity-aware embeddings. We examine a variety of applications and we thereby demonstrate the effectiveness of our PEM model. We also discuss important limitations of our work and stress caution when applying the PEM model to real-world scenarios.

Reinforcement learning (RL) techniques have been developed to optimize industrial cooling systems, offering substantial energy savings compared to traditional heuristic policies. A major challenge in industrial control involves learning behaviors that are feasible in the real world due to machinery constraints. For example, certain actions can only be executed every few hours while other actions can be taken more frequently. Without extensive reward engineering and experimentation, an RL agent may not learn realistic operation of machinery. To address this, we use hierarchical reinforcement learning with multiple agents that control subsets of actions according to their operation time scales. Our hierarchical approach achieves energy savings over existing baselines while maintaining constraints such as operating chillers within safe bounds in a simulated HVAC control environment.

The $h$-index is a metric used to measure the impact of a user in a publication setting, such as a member of a social network with many highly liked posts or a researcher in an academic domain with many highly cited publications. Specifically, the $h$-index of a user is the largest integer $h$ such that at least $h$ publications of the user have at least $h$ units of positive feedback. We design an algorithm that, given query access to the $n$ publications of a user and each publication's corresponding positive feedback number, outputs a $(1\pm \varepsilon)$-approximation of the $h$-index of this user with probability at least $1-δ$ in time \[ O(\frac{n \cdot \ln{(1/δ)}}{\varepsilon^2 \cdot h}), \] where $h$ is the actual $h$-index which is unknown to the algorithm a-priori. We then design a novel lower bound technique that allows us to prove that this bound is in fact asymptotically optimal for this problem in all parameters $n,h,\varepsilon,$ and $δ$. Our work is one of the first in sublinear time algorithms that addresses obtaining asymptotically optimal bounds, especially in terms of the error and confidence parameters. As such, we focus on designing novel techniques for this task. In particular, our lower bound technique seems quite general -- to showcase this, we also use our approach to prove an asymptotically optimal lower bound for the problem of estimating the number of triangles in a graph in sublinear time, which now is also optimal in the error and confidence parameters. This result improves upon prior lower bounds of Eden, Levi, Ron, and Seshadhri (FOCS'15) for this problem, as well as multiple follow-ups that extended this lower bound to other subgraph counting problems.

This paper studies the consequences of a human-initiated targeted attack on the national electric power system. We consider two kinds of attacks: ($i$) an attack by an adversary that uses a tactical weapon and destroys a large part of the grid, by physically targeting a large geographic region; ($ii$) a targeted attack by an adversary that takes out a small number of critical components in the network simultaneously. Our analysis uses ($i$) a realistic representation of the underlying power grid, including the topology, the control and protection components, ($ii$) a realistic representation of the targeted attack scenario, and ($iii$) a dynamic stability analysis, that goes beyond traditional work comprising structural and linear flow analysis. Such realistic analysis is expensive, but critical since it can capture cascading failures that result from transient instabilities introduced due to the attack. Our model acknowledges the presence of hidden failures in the protection systems resulting in relay misoperations. We analyze the extent of cascading outages for different levels of hidden failures. Our results show that: ($i$) the power grid is vulnerable to both these attacks, ($ii$) the tactical attack has significant social, economic and health damage but need not result in a regional cascade; on the contrary the targeted attack can cause significant cascade and lead to power outage over a large region. Our work shows the necessity to harden the power grid not just to cyber-attacks but also to physical attacks. Furthermore, we show that realistic representations and analysis can lead to fundamentally new insights that simplified models are unlikely to capture. Finally, the methods and results help us identify critical elements in the grid; the system can then be hardened in a more precise manner to reduce the vulnerabilities.

An independent, automated method of decoding and transcribing oral speech is known as automatic speech recognition (ASR). A typical ASR system extracts featured from audio recordings or streams and run one or more algorithms to map the features to corresponding texts. Numerous of research has been done in the field of speech signal processing in recent years. When given adequate resources, both conventional ASR and emerging end-to-end (E2E) speech recognition have produced promising results. However, for low-resource languages like Bengali, the current state of ASR lags behind, although the low resource state does not reflect upon the fact that this language is spoken by over 500 million people all over the world. Despite its popularity, there aren't many diverse open-source datasets available, which makes it difficult to conduct research on Bengali speech recognition systems. This paper is a part of the competition named `BUET CSE Fest DL Sprint'. The purpose of this paper is to improve the speech recognition performance of the Bengali language by adopting speech recognition technology on the E2E structure based on the transfer learning framework. The proposed method effectively models the Bengali language and achieves 3.819 score in `Levenshtein Mean Distance' on the test dataset of 7747 samples, when only 1000 samples of train dataset were used to train.

This work provides formal safety guarantees for control systems with disturbance. A disturbance observer-based robust safety-critical controller is proposed, that estimates the effect of the disturbance on safety and utilizes this estimate with control barrier functions to attain provably safe dynamic behavior. The observer error bound - which consists of transient and steady-state parts - is quantified, and the system is endowed with robustness against this error via the proposed controller. An adaptive cruise control problem is used as illustrative example through simulations including real disturbance data.