Search

Bitmap indexes are widely used for read-intensive analytical workloads because they are clustered and offer efficient reads with a small memory footprint. However, they are notoriously inefficient to update. As analytical applications are increasingly fused with transactional applications, leading to the emergence of hybrid transactional/analytical processing (HTAP), it is desirable that bitmap indexes support efficient concurrent real-time updates. In this paper, we propose Concurrent Updatable Bitmap indexing (CUBIT) that offers efficient real-time updates that scale with the number of CPU cores used and do not interfere with queries. Our design relies on three principles. First, we employ a horizontal bitwise representation of updated bits, which enables efficient atomic updates without locking entire bitvectors. Second, we propose a lightweight snapshotting mechanism that allows queries (including range queries) to run on separate snapshots and provides a wait-free progress guarantee. Third, we consolidate updates in a latch-free manner, providing a strong progress guarantee. Our evaluation shows that CUBIT offers 3x - 16x higher throughput and 3x - 220x lower latency than state-of-the-art updatable bitmap indexes. CUBIT's update-friendly nature widens the applicability of bitmap indexing. Experimenting with OLAP workloads with standard, batched updates shows that CUBIT overcomes the maintenance downtime and outperforms DuckDB by 1.2x - 2.7x on TPC-H. For HTAP workloads with real-time updates, CUBIT achieves 2x - 11x performance improvement over the state-of-the-art approaches.

Infectious diseases occur when pathogens from other individuals or animals infect a person, resulting in harm to both individuals and society as a whole. The outbreak of such diseases can pose a significant threat to human health. However, early detection and tracking of these outbreaks have the potential to reduce the mortality impact. To address these threats, public health authorities have endeavored to establish comprehensive mechanisms for collecting disease data. Many countries have implemented infectious disease surveillance systems, with the detection of epidemics being a primary objective. The clinical healthcare system, local/state health agencies, federal agencies, academic/professional groups, and collaborating governmental entities all play pivotal roles within this system. Moreover, nowadays, search engines and social media platforms can serve as valuable tools for monitoring disease trends. The Internet and social media have become significant platforms where users share information about their preferences and relationships. This real-time information can be harnessed to gauge the influence of ideas and societal opinions, making it highly useful across various domains and research areas, such as marketing campaigns, financial predictions, and public health, among others. This article provides a review of the existing standard methods developed by researchers for detecting outbreaks using time series data. These methods leverage various data sources, including conventional data sources and social media data or Internet data sources. The review particularly concentrates on works published within the timeframe of 2015 to 2022., Comment: 40 pages, 2 figures

In this paper, we address the challenge of heterogeneous data distributions in cross-silo federated learning by introducing a novel algorithm, which we term Cross-silo Robust Clustered Federated Learning (CS-RCFL). Our approach leverages the Wasserstein distance to construct ambiguity sets around each client's empirical distribution that capture possible distribution shifts in the local data, enabling evaluation of worst-case model performance. We then propose a model-agnostic integer fractional program to determine the optimal distributionally robust clustering of clients into coalitions so that possible biases in the local models caused by statistically heterogeneous client datasets are avoided, and analyze our method for linear and logistic regression models. Finally, we discuss a federated learning protocol that ensures the privacy of client distributions, a critical consideration, for instance, when clients are healthcare institutions. We evaluate our algorithm on synthetic and real-world healthcare data., Comment: 8 pages, 3 figures, Accepted to IEEE CDC 2024

In computer graphics (CG) education, the challenge of finding modern, versatile tools is significant, particularly when integrating both legacy and advanced technologies. Traditional frameworks, often reliant on solid, yet outdated APIs like OpenGL, limit the exploration of cutting-edge graphics techniques. To address this, we introduce pyGANDALF, a unique, lightweight, open-source CG framework built on three pillars: Entity-Component-System (ECS) architecture, Python programming, and WebGPU integration. This combination sets pyGANDALF apart by providing a streamlined ECS design with an editor layer, compatibility with WebGPU for state-of-the-art features like compute and ray tracing pipelines, and a programmer-friendly Python environment. The framework supports modern features, such as Physically Based Rendering (PBR) capabilities and integration with Universal Scene Description (USD) formats, making it suitable for both educational demonstrations and real-world applications. Evaluations by expert users confirmed that pyGANDALF effectively balances ease of use with advanced functionality, preparing students for contemporary CG development challenges., Comment: 8 pages, 6 Figures

Multicopter drones are becoming a key platform in several application domains, enabling precise on-the-spot sensing and/or actuation. We focus on the case where the drone must process the sensor data in order to decide, depending on the outcome, whether it needs to perform some additional action, e.g., more accurate sensing or some form of actuation. On the one hand, waiting for the computation to complete may waste time, if it turns out that no further action is needed. On the other hand, if the drone starts moving toward the next point of interest before the computation ends, it may need to return back to the previous point, if some action needs to be taken. In this paper, we propose a learning approach that enables the drone to take informed decisions about whether to wait for the result of the computation (or not), based on past experience gathered from previous missions. Through an extensive evaluation, we show that the proposed approach, when properly configured, outperforms several static policies, up to 25.8%, over a wide variety of different scenarios where the probability of some action being required at a given point of interest remains stable as well as for scenarios where this probability varies in time., Comment: 9 pages, 9 figures

Recent advances in Diffusion Models (DMs) have led to significant progress in visual synthesis and editing tasks, establishing them as a strong competitor to Generative Adversarial Networks (GANs). However, the latent space of DMs is not as well understood as that of GANs. Recent research has focused on unsupervised semantic discovery in the latent space of DMs by leveraging the bottleneck layer of the denoising network, which has been shown to exhibit properties of a semantic latent space. However, these approaches are limited to discovering global attributes. In this paper we address, the challenge of local image manipulation in DMs and introduce an unsupervised method to factorize the latent semantics learned by the denoising network of pre-trained DMs. Given an arbitrary image and defined regions of interest, we utilize the Jacobian of the denoising network to establish a relation between the regions of interest and their corresponding subspaces in the latent space. Furthermore, we disentangle the joint and individual components of these subspaces to identify latent directions that enable local image manipulation. Once discovered, these directions can be applied to different images to produce semantically consistent edits, making our method suitable for practical applications. Experimental results on various datasets demonstrate that our method can produce semantic edits that are more localized and have better fidelity compared to the state-of-the-art., Comment: Accepted at BMVC2024

Online conversations are particularly susceptible to derailment, which can manifest itself in the form of toxic communication patterns including disrespectful comments and abuse. Forecasting conversation derailment predicts signs of derailment in advance enabling proactive moderation of conversations. State-of-the-art approaches to conversation derailment forecasting sequentially encode conversations and use graph neural networks to model dialogue user dynamics. However, existing graph models are not able to capture complex conversational characteristics such as context propagation and emotional shifts. The use of common sense knowledge enables a model to capture such characteristics, thus improving performance. Following this approach, here we derive commonsense statements from a knowledge base of dialogue contextual information to enrich a graph neural network classification architecture. We fuse the multi-source information on utterance into capsules, which are used by a transformer-based forecaster to predict conversation derailment. Our model captures conversation dynamics and context propagation, outperforming the state-of-the-art models on the CGA and CMV benchmark datasets, Comment: arXiv admin note: substantial text overlap with arXiv:2306.12982; text overlap with arXiv:2106.01071 by other authors

Utilizing deep learning models to learn part-based representations holds significant potential for interpretable-by-design approaches, as these models incorporate latent causes obtained from feature representations through simple addition. However, training a part-based learning model presents challenges, particularly in enforcing non-negative constraints on the model's parameters, which can result in training difficulties such as instability and convergence issues. Moreover, applying such approaches in Deep Reinforcement Learning (RL) is even more demanding due to the inherent instabilities that impact many optimization methods. In this paper, we propose a non-negative training approach for actor models in RL, enabling the extraction of part-based representations that enhance interpretability while adhering to non-negative constraints. To this end, we employ a non-negative initialization technique, as well as a modified sign-preserving training method, which can ensure better gradient flow compared to existing approaches. We demonstrate the effectiveness of the proposed approach using the well-known Cartpole benchmark.

This paper introduces a novel integration of Large Language Models (LLMs) with Conformal Geometric Algebra (CGA) to revolutionize controllable 3D scene editing, particularly for object repositioning tasks, which traditionally requires intricate manual processes and specialized expertise. These conventional methods typically suffer from reliance on large training datasets or lack a formalized language for precise edits. Utilizing CGA as a robust formal language, our system, shenlong, precisely models spatial transformations necessary for accurate object repositioning. Leveraging the zero-shot learning capabilities of pre-trained LLMs, shenlong translates natural language instructions into CGA operations which are then applied to the scene, facilitating exact spatial transformations within 3D scenes without the need for specialized pre-training. Implemented in a realistic simulation environment, shenlong ensures compatibility with existing graphics pipelines. To accurately assess the impact of CGA, we benchmark against robust Euclidean Space baselines, evaluating both latency and accuracy. Comparative performance evaluations indicate that shenlong significantly reduces LLM response times by 16% and boosts success rates by 9.6% on average compared to the traditional methods. Notably, shenlong achieves a 100% perfect success rate in common practical queries, a benchmark where other systems fall short. These advancements underscore shenlong's potential to democratize 3D scene editing, enhancing accessibility and fostering innovation across sectors such as education, digital entertainment, and virtual reality., Comment: 17 pages, 8 figures

We present the observations and the results of our experience from a large number of flight hours in constructing, handling, testing, studying and interacting with our Antikythera Mechanism functional reconstruction models. We constructed our models according to the X Ray tomographies and without any use of modern stabilization parts (screws, nuts etc). Simple typical bronze material (not with special alloys) was used in the construction, while (most of) the gears were cut from simple bronze discs with thickness about 2mm and with triangular shaped teeth (not in modern involute shape). The bronze (not iron or steel) shafts have dimensions conforming to the original parts of the ancient artifact. The assembly of the parts was done following careful study of the Personal Constructional Characteristics and the Design Style of the ancient Craftsman, which are visible or even hidden on the Antikythera Mechanism Fragments. During the extensive use of our models, it was concluded that two important and mandatory indicators are missing from all current reconstructions of the Antikythera Mechanism. The absence of these two indicator dials makes it difficult to properly operate the Mechanism and their existence is necessary, in order for the Antikythera Mechanism to be considered as a complete and self contained operational time measuring device. The two procedures related to the (poorly) preserved remains were located on gear b1 and the lost Cover Disc of the b1 gear. The reconstruction of those missing parts was achieved by paying special attention to the Personal Constructional Characteristics of the ancient Craftsman. The extensive analysis of the Antikythera Mechanism calibrated scales leads to the understanding of the Mechanism as a luni solar time measuring device, as opposed to the notion that it was a mechanical planetarium presenting the hypothesized planetary motions and positions., Comment: 30 pages, 18 Figures

This paper introduces GA-Unity, the first Unity package specifically designed for seamless integration of Geometric Algebra (GA) into collaborative networked applications. Indeed, in such contexts, it has been demonstrated that using multivectors for interpolation between transmitted poses reduces runtime by 16% and bandwidth usage by an average of 50% compared to traditional representation forms (vectors and quaternions); we demonstrate that GA-Unity further enhances runtime performance. Tailored for 3D Conformal Geometric Algebra, GA-Unity also offers an intuitive interface within the Unity game engine, simplifying GA integration for researchers and programmers. By eliminating the need for users to develop GA functionalities from scratch, GA-Unity expedites GA experimentation and implementation processes. Its seamless integration enables easy representation of transformation properties using multivectors, facilitating deformations and interpolations without necessitating modifications to the rendering pipeline. Furthermore, its graphical interface establishes a GA playground for developers within the familiar confines of a modern game engine. In summary, GA-Unity represents a significant advancement in GA accessibility and usability, particularly in collaborative networked environments, empowering innovation and facilitating widespread adoption across various research and programming domains while upholding high-performance standards., Comment: 15 pages, 2 figures, paper refined and shortened

In this paper we present results from a qualitative field study on explainable AI (XAI) for lay users (n = 18) who were subjected to AI cyberattacks. The study was based on a custom-built smart heating application called Squid and was conducted over seven weeks in early 2023. Squid combined a smart radiator valve installed in participant homes with a web application that implemented an AI feature known as setpoint learning, which is commonly available in consumer smart thermostats. Development of Squid followed the XAI principle of interpretability-by-design where the AI feature was implemented using a simple glass-box machine learning model with the model subsequently exposed to users via the web interface (e.g. as interactive visualisations). AI attacks on users were simulated by injecting malicious training data and by manipulating data used for model predictions. Research data consisted of semi-structured interviews, researcher field notes, participant diaries, and application logs. In our analysis we reflect on the impact of XAI on user satisfaction and user comprehension as well as its use as a tool for diagnosing AI attacks. Our results show only limited engagement with XAI features and suggest that, for Squid users, common assumptions found in the XAI literature were not aligned to reality. On the positive side, users appear to have developed better mental models of the AI feature compared to previous work, and there is evidence that users did make some use of XAI as a diagnostic tool.

Barred galaxies constitute about two thirds of observed disc galaxies. Bars affect not only the mass distribution of gas and stars, but also that of the dark matter. An elongation of the inner dark matter halo is known as the halo bar. We aim to characterise the structure of the halo bars, with the goal of correlating them with the properties of the stellar bars. We use a suite of simulated galaxies with various bar strengths, including gas and star formation. We quantify strengths, shapes, and densities of these simulated stellar bars. We carry out numerical experiments with frozen and analytic potentials in order to understand the role played by a live responsive stellar bar. We find that the halo bar generally follows the trends of the disc bar. The strengths of the halo and stellar bars are tightly correlated. Stronger bars induce a slight increase of dark matter density within the inner halo. Numerical experiments show that a non-responsive frozen stellar bar would be capable of inducing a dark matter bar, but it would be weaker than the live case by a factor of roughly two., Comment: 18 pages, 9 figures, accepted for publication in Galaxies

Contemporary progress in the field of robotics, marked by improved efficiency and stability, has paved the way for the global adoption of surgical robotic systems (SRS). While these systems enhance surgeons' skills by offering a more accurate and less invasive approach to operations, they come at a considerable cost. Moreover, SRS components often involve heavy machinery, making the training process challenging due to limited access to such equipment. In this paper we introduce a cost-effective way to facilitate training for a simulator of a SRS via a portable, device-agnostic, ultra realistic simulation with hand tracking and feet tracking support. Error assessment is accessible in both real-time and offline, which enables the monitoring and tracking of users' performance. The VR application has been objectively evaluated by several untrained testers showcasing significant reduction in error metrics as the number of training sessions increases. This indicates that the proposed VR application denoted as VR Isle Academy operates efficiently, improving the robot - controlling skills of the testers in an intuitive and immersive way towards reducing the learning curve at minimal cost., Comment: 10 pages, 14 figures, Acknowledgement Section updated

The generative AI technology offers an increasing variety of tools for generating entirely synthetic images that are increasingly indistinguishable from real ones. Unlike methods that alter portions of an image, the creation of completely synthetic images presents a unique challenge and several Synthetic Image Detection (SID) methods have recently appeared to tackle it. Yet, there is often a large gap between experimental results on benchmark datasets and the performance of methods in the wild. To better address the evaluation needs of SID and help close this gap, this paper introduces a benchmarking framework that integrates several state-of-the-art SID models. Our selection of integrated models was based on the utilization of varied input features, and different network architectures, aiming to encompass a broad spectrum of techniques. The framework leverages recent datasets with a diverse set of generative models, high level of photo-realism and resolution, reflecting the rapid improvements in image synthesis technology. Additionally, the framework enables the study of how image transformations, common in assets shared online, such as JPEG compression, affect detection performance. SIDBench is available on https://github.com/mever-team/sidbench and is designed in a modular manner to enable easy inclusion of new datasets and SID models.

Malware sandboxes provide many benefits for security applications, but they are complex. These complexities can overwhelm new users in different research areas and make it difficult to select, configure, and use sandboxes. Even worse, incorrectly using sandboxes can have a negative impact on security applications. In this paper, we address this knowledge gap by systematizing 84 representative papers for using x86/64 malware sandboxes in the academic literature. We propose a novel framework to simplify sandbox components and organize the literature to derive practical guidelines for using sandboxes. We evaluate the proposed guidelines systematically using three common security applications and demonstrate that the choice of different sandboxes can significantly impact the results. Specifically, our results show that the proposed guidelines improve the sandbox observable activities by at least 1.6x and up to 11.3x. Furthermore, we observe a roughly 25% improvement in accuracy, precision, and recall when using the guidelines to help with a malware family classification task. We conclude by affirming that there is no "silver bullet" sandbox deployment that generalizes, and we recommend that users apply our framework to define a scope for their analysis, a threat model, and derive context about how the sandbox artifacts will influence their intended use case. Finally, it is important that users document their experiment, limitations, and potential solutions for reproducibility

In this work, we present a study on ways that tracking algorithms can be improved with machine learning (ML). We base this study on the line segment tracking (LST) algorithm that we have designed to be naturally parallelized and vectorized in order to efficiently run on modern processors. LST has been developed specifically for the CMS Experiment at the LHC, towards the High Luminosity LHC (HL-LHC) upgrade. Moreover, we have already shown excellent efficiency and performance results as we iteratively improve LST, leveraging a full simulation of the CMS detector. At the same time, promising deep-learning-based tracking algorithms, such as Graph Neural Networks (GNNs), are being pioneered on the simplified TrackML dataset. These results suggest that parts of LST could be improved or replaced by ML. Thus, a thorough, step-by-step investigation of exactly how and where ML can be utilized, while still meeting realistic HL-LHC performance and efficiency constraints, is implemented as follows. First, a lightweight neural network is used to replace and improve upon explicitly defined track quality selections. This neural network is shown to be highly efficient and robust to displaced tracks while having little-to-no impact on the runtime of LST. These results clearly establish that ML can be used to improve LST without penalty. Next, exploratory studies of GNN track-building algorithms are described. In particular, low-level track objects from LST are considered as nodes in a graph, where edges represent higher-level objects or even entire track candidates. Then, an edge-classifier GNN is trained, and the efficiency of the resultant edge scores is compared with that of the existing LST track quality selections. These GNN studies provide insights into the practicality and performance of using more ambitious and complex ML algorithms for HL-LHC tracking at the CMS Experiment., Comment: Connecting the Dots Workshop (CTD 2023)

This study introduces an novel approach to the design of low-profile superdirective antenna arrays, employing parasitic elements. The proposed design concept was verified through the analysis of a two-element antenna array consisting of strip dipoles, operating at a frequency of $3.5$ GHz within the sub-6 5G frequency band. The antenna array was optimized for realized gain, a key antenna parameter considering both ohmic and return losses. The design parameters included the lengths and widths of the strip dipoles, along with a reactance load connected to the parasitic element. The driven element was excited by a sinusoidal voltage signal with a magnitude of $1$ V, eliminating the need for amplifiers, attenuators, phase shifters, or impedance matching networks. Results demonstrated that this design concept enables the achievement of superdirectivity, with inter-element distances as small as $0.1\lambda$, resulting in low-profile, compact, high directional antenna systems.

This study presents a superdirective antenna array specifically designed for the sub-6 5G generation frequency range, incorporating pioneering Huygens antenna elements. The optimized structure achieves a realized gain that surpasses Harrington's well-known maximum theoretical limit for antenna directivity, effectively addressing practical concerns related to ohmic and return losses. Additionally, the compact size of the proposed antenna array, as opposed to a uniform linear array, offers the potential to fabricate highly radiation-efficient, high-directivity antennas in a compact form., Comment: 09/02/2024

In a modern DBMS, working memory is frequently the limiting factor when processing in-memory analytic query operations such as joins, sorting, and aggregation. Existing resource estimation approaches for a DBMS estimate the resource consumption of a query by computing an estimate of each individual database operator in the query execution plan. Such an approach is slow and error-prone as it relies upon simplifying assumptions, such as uniformity and independence of the underlying data. Additionally, the existing approach focuses on individual queries separately and does not factor in other queries in the workload that may be executed concurrently. In this research, we are interested in query performance optimization under concurrent execution of a batch of queries (a workload). Specifically, we focus on predicting the memory demand for a workload rather than providing separate estimates for each query within it. We introduce the problem of workload memory prediction and formalize it as a distribution regression problem. We propose Learned Workload Memory Prediction (LearnedWMP) to improve and simplify estimating the working memory demands of workloads. Through a comprehensive experimental evaluation, we show that LearnedWMP reduces the memory estimation error of the state-of-the-practice method by up to 47.6%. Compared to an alternative single-query model, during training and inferencing, the LearnedWMP model and its variants were 3x to 10x faster. Moreover, LearnedWMP-based models were at least 50% smaller in most cases. Overall, the results demonstrate the advantages of the LearnedWMP approach and its potential for a broader impact on query performance optimization.

Modern GPU applications, such as machine learning (ML), can only partially utilize GPUs, leading to GPU underutilization in cloud environments. Sharing GPUs across multiple applications from different tenants can improve resource utilization and consequently cost, energy, and power efficiency. However, GPU sharing creates memory safety concerns because kernels must share a single GPU address space. Existing spatial-sharing mechanisms either lack fault isolation for memory accesses or require static partitioning, which leads to limited deployability or low utilization. In this paper, we present Guardian, a PTX-level bounds checking approach that provides memory isolation and supports dynamic GPU spatial-sharing. Guardian relies on three mechanisms: (1) It divides the common GPU address space into separate partitions for different applications. (2) It intercepts and checks all GPU related calls at the lowest level, fencing erroneous operations. (3) It instruments all GPU kernels at the PTX level -- available in closed GPU libraries -- fencing all kernel memory accesses outside application memory bounds. Guardian's approach is transparent to applications and supports real-life frameworks, such as Caffe and PyTorch, that issue billions of GPU kernels. Our evaluation shows that Guardian's overhead compared to native for such frameworks is between 4% - 12% and on average 9%.

Understanding and predicting the emotional trajectory in multi-party multi-turn conversations is of great significance. Such information can be used, for example, to generate empathetic response in human-machine interaction or to inform models of pre-emptive toxicity detection. In this work, we introduce the novel problem of Predicting Emotions in Conversations (PEC) for the next turn (n+1), given combinations of textual and/or emotion input up to turn n. We systematically approach the problem by modeling three dimensions inherently connected to evoked emotions in dialogues, including (i) sequence modeling, (ii) self-dependency modeling, and (iii) recency modeling. These modeling dimensions are then incorporated into two deep neural network architectures, a sequence model and a graph convolutional network model. The former is designed to capture the sequence of utterances in a dialogue, while the latter captures the sequence of utterances and the network formation of multi-party dialogues. We perform a comprehensive empirical evaluation of the various proposed models for addressing the PEC problem. The results indicate (i) the importance of the self-dependency and recency model dimensions for the prediction task, (ii) the quality of simpler sequence models in short dialogues, (iii) the importance of the graph neural models in improving the predictions in long dialogues.

Replicated state machines (RSMs) cannot effectively communicate today as there is no formal framework or efficient protocol to do so. To address this issue, we introduce a new primitive, the Cross-Cluster Consistent Broadcast (C3B) and present PICSOU, a practical C3B implementation. PICSOU draws inspiration from networking and TCP to allow two RSMs to communicate with constant metadata overhead in the failure-free case and minimal number of message resends in the case of failures. PICSOU is flexible and allows both crash fault-tolerant and byzantine fault-tolerant protocols to communicate. At the heart of PICSOU's good performance and generality lies a novel technique we call QUACKs (quorum acknowledgements) that allow nodes in each RSM to precisely determine when messages have definitely been received, or definitely been lost. Our results are promising: we obtain up to 24x better performance than existing all-to-all solutions.

Collecting data, extracting value, and combining insights from relational and context-rich multi-modal sources in data processing pipelines presents a challenge for traditional relational DBMS. While relational operators allow declarative and optimizable query specification, they are limited to data transformations unsuitable for capturing or analyzing context. On the other hand, representation learning models can map context-rich data into embeddings, allowing machine-automated context processing but requiring imperative data transformation integration with the analytical query. To bridge this dichotomy, we present a context-enhanced relational join and introduce an embedding operator composable with relational operators. This enables hybrid relational and context-rich vector data processing, with algebraic equivalences compatible with relational algebra and corresponding logical and physical optimizations. We investigate model-operator interaction with vector data processing and study the characteristics of the E-join operator. Using an example of string embeddings, we demonstrate enabling hybrid context-enhanced processing on relational join operators with vector embeddings. The importance of holistic optimization, from logical to physical, is demonstrated in an order of magnitude execution time improvement.

Log-Structured Merge trees (LSM trees) are increasingly used as part of the storage engine behind several data systems, and are frequently deployed in the cloud. As the number of applications relying on LSM-based storage backends increases, the problem of performance tuning of LSM trees receives increasing attention. We consider both nominal tunings - where workload and execution environment are accurately known a priori - and robust tunings - which consider uncertainty in the workload knowledge. This type of workload uncertainty is common in modern applications, notably in shared infrastructure environments like the public cloud. To address this problem, we introduce ENDURE, a new paradigm for tuning LSM trees in the presence of workload uncertainty. Specifically, we focus on the impact of the choice of compaction policy, size ratio, and memory allocation on the overall performance. ENDURE considers a robust formulation of the throughput maximization problem and recommends a tuning that offers near-optimal throughput when the executed workload is not the same, instead in a neighborhood of the expected workload. Additionally, we explore the robustness of flexible LSM designs by proposing a new unified design called K-LSM that encompasses existing designs. We deploy our robust tuning system, ENDURE, on a state-of-the-art key-value store, RocksDB, and demonstrate throughput improvements of up to 5x in the presence of uncertainty. Our results indicate that the tunings obtained by ENDURE are more robust than tunings obtained under our expanded LSM design space. This indicates that robustness may not be inherent to a design, instead, it is an outcome of a tuning process that explicitly accounts for uncertainty., Comment: 25 pages, 19 figures, VLDB-J. arXiv admin note: substantial text overlap with arXiv:2110.13801

The healthcare sector is increasingly vulnerable to cyberattacks due to its growing digitalization. Patient data, including medical records and financial information, are at risk, potentially leading to identity theft and patient safety concerns. The European Union and other organizations identify key areas for healthcare system improvement, yet the industry still grapples with inadequate security practices. In response, the HEIR project offers a comprehensive cybersecurity approach, promoting security features from various regulatory frameworks and introducing tools such as the Secure Healthcare Framework and Risk Assessment for Medical Applications (RAMA). These measures aim to enhance digital health security and protect sensitive patient data while facilitating secure data access and privacy-aware techniques. In a rapidly evolving threat landscape, HEIR presents a promising framework for healthcare cybersecurity., Comment: 70 pages, 48 figures

Storage-based joins are still commonly used today because the memory budget does not always scale with the data size. One of the many join algorithms developed that has been widely deployed and proven to be efficient is the Hybrid Hash Join (HHJ), which is designed to exploit any available memory to maximize the data that is joined directly in memory. However, HHJ cannot fully exploit detailed knowledge of the join attribute correlation distribution. In this paper, we show that given a correlation skew in the join attributes, HHJ partitions data in a suboptimal way. To do that, we derive the optimal partitioning using a new cost-based analysis of partitioning-based joins that is tailored for primary key - foreign key (PK-FK) joins, one of the most common join types. This optimal partitioning strategy has a high memory cost, thus, we further derive an approximate algorithm that has tunable memory cost and leads to near-optimal results. Our algorithm, termed NOCAP (Near-Optimal Correlation-Aware Partitioning) join, outperforms the state-of-the-art for skewed correlations by up to $30\%$, and the textbook Grace Hash Join by up to $4\times$. Further, for a limited memory budget, NOCAP outperforms HHJ by up to $10\%$, even for uniform correlation. Overall, NOCAP dominates state-of-the-art algorithms and mimics the best algorithm for a memory budget varying from below $\sqrt{\|\text{relation}\|}$ to more than $\|\text{relation}\|$.

Neuromorphic photonic accelerators are becoming increasingly popular, since they can significantly improve computation speed and energy efficiency, leading to femtojoule per MAC efficiency. However, deploying existing DL models on such platforms is not trivial, since a great range of photonic neural network architectures relies on incoherent setups and power addition operational schemes that cannot natively represent negative quantities. This results in additional hardware complexity that increases cost and reduces energy efficiency. To overcome this, we can train non-negative neural networks and potentially exploit the full range of incoherent neuromorphic photonic capabilities. However, existing approaches cannot achieve the same level of accuracy as their regular counterparts, due to training difficulties, as also recent evidence suggests. To this end, we introduce a methodology to obtain the non-negative isomorphic equivalents of regular neural networks that meet requirements of neuromorphic hardware, overcoming the aforementioned limitations. Furthermore, we also introduce a sign-preserving optimization approach that enables training of such isomorphic networks in a non-negative manner.

In this work, we investigate the personalization of text-to-music diffusion models in a few-shot setting. Motivated by recent advances in the computer vision domain, we are the first to explore the combination of pre-trained text-to-audio diffusers with two established personalization methods. We experiment with the effect of audio-specific data augmentation on the overall system performance and assess different training strategies. For evaluation, we construct a novel dataset with prompts and music clips. We consider both embedding-based and music-specific metrics for quantitative evaluation, as well as a user study for qualitative evaluation. Our analysis shows that similarity metrics are in accordance with user preferences and that current personalization approaches tend to learn rhythmic music constructs more easily than melody. The code, dataset, and example material of this study are open to the research community., Comment: Submitted to ICASSP 2024, Examples at https://zelaki.github.io/

In this study, we investigate and fabricate a superdirective antenna array composed of strip dipole elements operating at a frequency of 3.5 GHz. The spacing, dimensions, and phase difference of the elements are optimized to achieve a super realized gain antenna with a theoretical efficiency of 98.8% and computed efficiency of 99.3%. By employing an element spacing of 0.2$\lambda$, the end-fire antenna array demonstrates a maximum theoretical realized gain of 6.4 dBi, and a maximum computed realized gain of 6.3 dBi. Significantly, our proposed superdirective antenna array distinguishes itself from existing approaches by achieving high directivity, high radiation efficiency, and impedance matching to 50 $\Omega$ solely through careful adjustments in the strip dimensions and the inter-element phase. This eliminates the need for additional impedance matching networks, amplifiers, or attenuators.

The latest cancer statistics indicate a decrease in cancer-related mortality. However, due to the growing and ageing population, the absolute number of people living with cancer is set to keep increasing. This paper presents ASCAPE, an open AI infrastructure that takes advantage of the recent advances in Artificial Intelligence (AI) and Machine Learning (ML) to support cancer patients quality of life (QoL). With ASCAPE health stakeholders (e.g. hospitals) can locally process their private medical data and then share the produced knowledge (ML models) through the open AI infrastructure.

Financial portfolio management describes the task of distributing funds and conducting trading operations on a set of financial assets, such as stocks, index funds, foreign exchange or cryptocurrencies, aiming to maximize the profit while minimizing the loss incurred by said operations. Deep Learning (DL) methods have been consistently excelling at various tasks and automated financial trading is one of the most complex one of those. This paper aims to provide insight into various DL methods for financial trading, under both the supervised and reinforcement learning schemes. At the same time, taking into consideration sentiment information regarding the traded assets, we discuss and demonstrate their usefulness through corresponding research studies. Finally, we discuss commonly found problems in training such financial agents and equip the reader with the necessary knowledge to avoid these problems and apply the discussed methods in practice.

Even nowadays, where Deep Learning (DL) has achieved state-of-the-art performance in a wide range of research domains, accelerating training and building robust DL models remains a challenging task. To this end, generations of researchers have pursued to develop robust methods for training DL architectures that can be less sensitive to weight distributions, model architectures and loss landscapes. However, such methods are limited to adaptive learning rate optimizers, initialization schemes, and clipping gradients without investigating the fundamental rule of parameters update. Although multiplicative updates have contributed significantly to the early development of machine learning and hold strong theoretical claims, to best of our knowledge, this is the first work that investigate them in context of DL training acceleration and robustness. In this work, we propose an optimization framework that fits to a wide range of optimization algorithms and enables one to apply alternative update rules. To this end, we propose a novel multiplicative update rule and we extend their capabilities by combining it with a traditional additive update term, under a novel hybrid update method. We claim that the proposed framework accelerates training, while leading to more robust models in contrast to traditionally used additive update rule and we experimentally demonstrate their effectiveness in a wide range of task and optimization methods. Such tasks ranging from convex and non-convex optimization to difficult image classification benchmarks applying a wide range of traditionally used optimization methods and Deep Neural Network (DNN) architectures.

Online conversations are particularly susceptible to derailment, which can manifest itself in the form of toxic communication patterns like disrespectful comments or verbal abuse. Forecasting conversation derailment predicts signs of derailment in advance enabling proactive moderation of conversations. Current state-of-the-art approaches to address this problem rely on sequence models that treat dialogues as text streams. We propose a novel model based on a graph convolutional neural network that considers dialogue user dynamics and the influence of public perception on conversation utterances. Through empirical evaluation, we show that our model effectively captures conversation dynamics and outperforms the state-of-the-art models on the CGA and CMV benchmark datasets by 1.5\% and 1.7\%, respectively., Comment: WOAH, ACL

This work presents the introduction of UniSG^GA, a novel integrated scenegraph structure, that to incorporates behavior and geometry data on a 3D scene. It is specifically designed to seamlessly integrate Graph Neural Networks (GNNs) and address the challenges associated with transforming a 3D scenegraph (3D-SG) during generative tasks. To effectively capture and preserve the topological relationships between objects in a simplified way, within the graph representation, we propose UniSG^GA, that seamlessly integrates Geometric Algebra (GA) forms. This novel approach enhances the overall performance and capability of GNNs in handling generative and predictive tasks, opening up new possibilities and aiming to lay the foundation for further exploration and development of graph-based generative AI models that can effectively incorporate behavior data for enhanced scene generation and synthesis., Comment: 7 pages, 5 figures, A version of this paper was submitted to the ENGAGE workshop of CGI 2023