arXiv Computer Science @arxiv_cs@qoto.org

Ionic Liquids (ILs) provide a promising solution for CO$_2$ capture and storage to mitigate global warming. However, identifying and designing the high-capacity IL from the giant chemical space requires expensive, and exhaustive simulations and experiments. Machine learning (ML) can accelerate the process of searching for desirable ionic molecules through accurate and efficient property predictions in a data-driven manner. But existing descriptors and ML models for the ionic molecule suffer from the inefficient adaptation of molecular graph structure. Besides, few works have investigated the explainability of ML models to help understand the learned features that can guide the design of efficient ionic molecules. In this work, we develop both fingerprint-based ML models and Graph Neural Networks (GNNs) to predict the CO$_2$ absorption in ILs. Fingerprint works on graph structure at the feature extraction stage, while GNNs directly handle molecule structure in both the feature extraction and model prediction stage. We show that our method outperforms previous ML models by reaching a high accuracy (MAE of 0.0137, $R^2$ of 0.9884). Furthermore, we take the advantage of GNNs feature representation and develop a substructure-based explanation method that provides insight into how each chemical fragments within IL molecules contribute to the CO$_2$ absorption prediction of ML models. We also show that our explanation result agrees with some ground truth from the theoretical reaction mechanism of CO$_2$ absorption in ILs, which can advise on the design of novel and efficient functional ILs in the future.

Human perception, memory and decision-making are impacted by tens of cognitive biases and heuristics that influence our actions and decisions. Despite the pervasiveness of such biases, they are generally not leveraged by today's Artificial Intelligence (AI) systems that model human behavior and interact with humans. In this theoretical paper, we claim that the future of human-machine collaboration will entail the development of AI systems that model, understand and possibly replicate human cognitive biases. We propose the need for a research agenda on the interplay between human cognitive biases and Artificial Intelligence. We categorize existing cognitive biases from the perspective of AI systems, identify three broad areas of interest and outline research directions for the design of AI systems that have a better understanding of our own biases.

Simulators driven by deep learning are gaining popularity as a tool for efficiently emulating accurate but expensive numerical simulators. Successful applications of such neural simulators can be found in the domains of physics, chemistry, and structural biology, amongst others. Likewise, a neural simulator for cellular dynamics can augment lab experiments and traditional computational methods to enhance our understanding of a cell's interaction with its physical environment. In this work, we propose an autoregressive probabilistic model that can reproduce spatiotemporal dynamics of single cell migration, traditionally simulated with the Cellular Potts model. We observe that standard single-step training methods do not only lead to inconsistent rollout stability, but also fail to accurately capture the stochastic aspects of the dynamics, and we propose training strategies to mitigate these issues. Our evaluation on two proof-of-concept experimental scenarios shows that neural methods have the potential to faithfully simulate stochastic cellular dynamics at least an order of magnitude faster than a state-of-the-art implementation of the Cellular Potts model.

Neural networks are configured by choosing an architecture and hyperparameter values; doing so often involves expert intuition and hand-tuning to find a configuration that extrapolates well without overfitting. This paper considers automatic methods for configuring a neural network that extrapolates in time for the domain of visible and near-infrared (VNIR) spectroscopy. In particular, we study the effect of (a) selecting samples for validating configurations and (b) using ensembles. Most of the time, models are built of the past to predict the future. To encourage the neural network model to extrapolate, we consider validating model configurations on samples that are shifted in time similar to the test set. We experiment with three validation set choices: (1) a random sample of 1/3 of non-test data (the technique used in previous work), (2) using the latest 1/3 (sorted by time), and (3) using a semantically meaningful subset of the data. Hyperparameter optimization relies on the validation set to estimate test-set error, but neural network variance obfuscates the true error value. Ensemble averaging - computing the average across many neural networks - can reduce the variance of prediction errors. To test these methods, we do a comprehensive study of a held-out 2018 harvest season of mango fruit given VNIR spectra from 3 prior years. We find that ensembling improves the state-of-the-art model's variance and accuracy. Furthermore, hyperparameter optimization experiments - with and without ensemble averaging and with each validation set choice - show that when ensembling is combined with using the latest 1/3 of samples as the validation set, a neural network configuration is found automatically that is on par with the state-of-the-art.

The photon-counting detector (PCD) based spectral computed tomography attracts much more attentions since it has the capability to provide more accurate identification and quantitative analysis for biomedical materials. The limited number of photons within narrow energy-bin leads to low signal-noise ratio data. The existing supervised deep reconstruction networks for CT reconstruction are difficult to address these challenges. In this paper, we propose an iterative deep reconstruction network to synergize model and data priors into a unified framework, named as Spectral2Spectral. Our Spectral2Spectral employs an unsupervised deep training strategy to obtain high-quality images from noisy data with an end-to-end fashion. The structural similarity prior within image-spectral domain is refined as a regularization term to further constrain the network training. The weights of neural network are automatically updated to capture image features and structures with iterative process. Three large-scale preclinical datasets experiments demonstrate that the Spectral2spectral reconstruct better image quality than other state-of-the-art methods.

The impact performance of the wheel during wheel development must be ensured through a wheel impact test for vehicle safety. However, manufacturing and testing a real wheel take a significant amount of time and money because developing an optimal wheel design requires numerous iterative processes of modifying the wheel design and verifying the safety performance. Accordingly, the actual wheel impact test has been replaced by computer simulations, such as Finite Element Analysis (FEA), but it still requires high computational costs for modeling and analysis. Moreover, FEA experts are needed. This study presents an aluminum road wheel impact performance prediction model based on deep learning that replaces the computationally expensive and time-consuming 3D FEA. For this purpose, 2D disk-view wheel image data, 3D wheel voxel data, and barrier mass value used for wheel impact test are utilized as the inputs to predict the magnitude of maximum von Mises stress, corresponding location, and the stress distribution of 2D disk-view. The wheel impact performance prediction model can replace the impact test in the early wheel development stage by predicting the impact performance in real time and can be used without domain knowledge. The time required for the wheel development process can be shortened through this mechanism.

We present a robust system for state estimation that fuses measurements from multiple lidars and inertial sensors with GNSS data. To initiate the method, we use the prior GNSS pose information. We then perform incremental motion in real-time, which produces robust motion estimates in a global frame by fusing lidar and IMU signals with GNSS translation components using a factor graph framework. We also propose methods to account for signal loss with a novel synchronization and fusion mechanism. To validate our approach extensive tests were carried out on data collected using Scania test vehicles (5 sequences for a total of ~ 7 Km). From our evaluations, we show an average improvement of 61% in relative translation and 42% rotational error compared to a state-of-the-art estimator fusing a single lidar/inertial sensor pair.

We show how the Weil pairing can be used to evaluate the assigned characters of an imaginary quadratic order $\mathcal{O}$ in an unknown ideal class $[\mathfrak{a}] \in \mathrm{Cl}(\mathcal{O})$ that connects two given $\mathcal{O}$-oriented elliptic curves $(E, ι)$ and $(E', ι') = [\mathfrak{a}](E, ι)$. When specialized to ordinary elliptic curves over finite fields, our method is conceptually simpler and often somewhat faster than a recent approach due to Castryck, Sotáková and Vercauteren, who rely on the Tate pairing instead. The main implication of our work is that it breaks the decisional Diffie-Hellman problem for practically all oriented elliptic curves that are acted upon by an even-order class group. It can also be used to better handle the worst cases in Wesolowski's recent reduction from the vectorization problem for oriented elliptic curves to the endomorphism ring problem, leading to a method that always works in sub-exponential time.

Synchronous updates may compromise the efficiency of cross-device federated learning once the number of active clients increases. The \textit{FedBuff} algorithm (Nguyen et al., 2022) alleviates this problem by allowing asynchronous updates (staleness), which enhances the scalability of training while preserving privacy via secure aggregation. We revisit the \textit{FedBuff} algorithm for asynchronous federated learning and extend the existing analysis by removing the boundedness assumptions from the gradient norm. This paper presents a theoretical analysis of the convergence rate of this algorithm when heterogeneity in data, batch size, and delay are considered.

This paper explores continuous-time control synthesis for target-driven navigation to satisfy complex high-level tasks expressed as linear temporal logic (LTL). We propose a model-free framework using deep reinforcement learning (DRL) where the underlying dynamic system is unknown (an opaque box). Unlike prior work, this paper considers scenarios where the given LTL specification might be infeasible and therefore cannot be accomplished globally. Instead of modifying the given LTL formula, we provide a general DRL-based approach to satisfy it with minimal violation. %\mminline{Need to decide if we're comfortable calling these "guarantees" due to the stochastic policy. I'm not repeating this comment everywhere that says "guarantees" but there are multiple places.} To do this, we transform a previously multi-objective DRL problem, which requires simultaneous automata satisfaction and minimum violation cost, into a single objective. By guiding the DRL agent with a sampling-based path planning algorithm for the potentially infeasible LTL task, the proposed approach mitigates the myopic tendencies of DRL, which are often an issue when learning general LTL tasks that can have long or infinite horizons. This is achieved by decomposing an infeasible LTL formula into several reach-avoid sub-tasks with shorter horizons, which can be trained in a modular DRL architecture. Furthermore, we overcome the challenge of the exploration process for DRL in complex and cluttered environments by using path planners to design rewards that are dense in the configuration space. The benefits of the presented approach are demonstrated through testing on various complex nonlinear systems and compared with state-of-the-art baselines. The Video demonstration can be found on YouTube Channel:\url{https://youtu.be/jBhx6Nv224E}.

A novel geometrically exact model of the spatially curved Bernoulli-Euler beam is developed. The formulation utilizes the Frenet-Serret frame as the reference for updating the orientation of a cross section. The weak form is consistently derived and linearized, including the contributions from kinematic constraints and configuration-dependent load. The nonlinear terms with respect to the cross-sectional coordinates are strictly considered, and the obtained constitutive model is scrutinized. The main features of the formulation are invariance with respect to the rigid-body motion, path-independence, and improved accuracy for strongly curved beams. A new reduced beam model is conceived as a special case, by omitting the rotational DOF. Although rotation-free, the reduced model includes the part of the torsional stiffness that is related to the torsion of the beam axis. This allows simulation of examples where the angle between material axes and Frenet-Serret frame is small. The applicability of the obtained isogeometric finite element is verified via a set of standard academic benchmark examples. The formulation is able to accurately model strongly curved Bernoulli-Euler beams that have well-defined Frenet-Serret frames.

In order to achieve a virtual certification process and robust designs for turbomachinery, the uncertainty bounds for computational fluid dynamics have to be known. The formulation of turbulence closure models implies a major source of the overall uncertainty of Reynold-averaged Navier Stokes simulations. We discuss the common practice of applying a physics constrained eigenspace perturbation of the Reynolds stress tensor in order to account for the model form uncertainty of turbulence models. Since the basic methodology often leads to generous uncertainty estimates, we extend a recent approach of adding a machine learning strategy. The application of a data-driven method is motivated by striving for the detection of flow regions, which are prone to suffer from a lack of turbulence model prediction accuracy. In this way any user input related to choosing the degree of uncertainty is supposed to become obsolete. This work especially investigates an approach, which tries to determine an a priori estimation of prediction confidence, when there is no accurate data available to judge the prediction. The flow around the NACA 4412 airfoil at near-stall conditions serves to demonstrate the successful application of the data-driven eigenspace perturbation framework. We especially highlight the objectives and limitations of the underlying methodology finally.

This paper proposes a methodology for architecting microstructures with extremal stiffness, yield, and buckling strength using topology optimization. The optimized microstructures reveal an interesting transition from simple lattice like structures for yield-dominated situations to hierarchical lattice structures for buckling-dominated situations. The transition from simple to hierarchical is governed by the relative yield strength of the constituent base material as well as the volume fraction. The overall performances of the proposed microstructures indicate that optimal strength is determined by the buckling strength at low volume fractions and yield strength at high volume fractions, regardless of the base material's relative yield strength.

The aim of this work is to present a series of concrete examples which illustrate how the computer algebra system Cadabra can be used to manipulate expressions appearing in General Relativity and other gravitational theories. We highlight the way in which Cadabra's philosophy differs from other systems with related functionality. The use of various new built-in packages is discussed, and we show how such packages can also be created by end-users directly using the notebook interface. The current paper focuses on fairly generic applications in gravitational theories, including the use of differential forms, the derivation of field equations and the construction of their solutions. A follow-up paper discusses more specific applications related to the analysis of gravitational waves.

Electron cryo-microscopy (cryo-EM) produces three-dimensional (3D) maps of the electrostatic potential of biological macromolecules, including proteins. At sufficient resolution, the cryo-EM maps, along with some knowledge about the imaged molecules, allow de novo atomic modelling. Typically, this is done through a laborious manual process. Recent advances in machine learning applications to protein structure prediction show potential for automating this process. Taking inspiration from these techniques, we have built ModelAngelo for automated model building of proteins in cryo-EM maps. ModelAngelo first uses a residual convolutional neural network (CNN) to initialize a graph representation with nodes assigned to individual amino acids of the proteins in the map and edges representing the protein chain. The graph is then refined with a graph neural network (GNN) that combines the cryo-EM data, the amino acid sequence data and prior knowledge about protein geometries. The GNN refines the geometry of the protein chain and classifies the amino acids for each of its nodes. The final graph is post-processed with a hidden Markov model (HMM) search to map each protein chain to entries in a user provided sequence file. Application to 28 test cases shows that ModelAngelo outperforms the state-of-the-art and approximates manual building for cryo-EM maps with resolutions better than 3.5 Å.

Computer Algebra Systems (CASs) like Cadabra Software play a prominent role in a wide range of research activities in physics and related fields. We show how Cadabra language is easily implemented in the well established Python programming framework, gaining excellent flexibility and customization to address the issue of tensor perturbations in General Relativity. We obtain a performing algorithm to decompose tensorial quantities up to any perturbative order of the metric. The features of our code are tested by discussing some concrete computational issues in research activities related to first/higher-order gravitational waves.

Signature-based malware detectors have proven to be insufficient as even a small change in malignant executable code can bypass these signature-based detectors. Many machine learning-based models have been proposed to efficiently detect a wide variety of malware. Many of these models are found to be susceptible to adversarial attacks - attacks that work by generating intentionally designed inputs that can force these models to misclassify. Our work aims to explore vulnerabilities in the current state of the art malware detectors to adversarial attacks. We train a Transformers-based malware detector, carry out adversarial attacks resulting in a misclassification rate of 23.9% and propose defenses that reduce this misclassification rate to half. An implementation of our work can be found at https://github.com/yashjakhotiya/Adversarial-Attacks-On-Transformers.

We introduce a fast direct solver for variable-coefficient elliptic partial differential equations on surfaces based on the hierarchical Poincaré-Steklov method. The method takes as input an unstructured, high-order quadrilateral mesh of a surface and discretizes surface differential operators on each element using a high-order spectral collocation scheme. Elemental solution operators and Dirichlet-to-Neumann maps tangent to the surface are precomputed and merged in a pairwise fashion to yield a hierarchy of solution operators that may be applied in $\mathcal{O}(N \log N)$ operations for a mesh with $N$ elements. The resulting fast direct solver may be used to accelerate high-order implicit time-stepping schemes, as the precomputed operators can be reused for fast elliptic solves on surfaces. On a standard laptop, precomputation for a 12th-order surface mesh with over 1 million degrees of freedom takes 17 seconds, while subsequent solves take only 0.25 seconds. We apply the method to a range of problems on both smooth surfaces and surfaces with sharp corners and edges, including the static Laplace-Beltrami problem, the Hodge decomposition of a tangential vector field, and some time-dependent nonlinear reaction-diffusion systems.

While standard bandit algorithms sometimes incur high regret, their performance can be greatly improved by "warm starting" with historical data. Unfortunately, how best to incorporate historical data is unclear: naively initializing reward estimates using all historical samples can suffer from spurious data and imbalanced data coverage, leading to computational and storage issues - particularly in continuous action spaces. We address these two challenges by proposing Artificial Replay, a meta-algorithm for incorporating historical data into any arbitrary base bandit algorithm. Artificial Replay uses only a subset of the historical data as needed to reduce computation and storage. We show that for a broad class of base algorithms that satisfy independence of irrelevant data (IIData), a novel property that we introduce, our method achieves equal regret as a full warm-start approach while potentially using only a fraction of the historical data. We complement these theoretical results with a case study of $K$-armed and continuous combinatorial bandit algorithms, including on a green security domain using real poaching data, to show the practical benefits of Artificial Replay in achieving optimal regret alongside low computational and storage costs.

We extend earlier work on the design of convolutional code-specific CRC codes to $Q$-ary alphabets, with an eye toward $Q$-ary orthogonal signaling. Starting with distance-spectrum optimal, zero-terminated, $Q$-ary convolutional codes, we design $Q$-ary CRC codes so that the CRC/convolutional concatenation is distance-spectrum optimal. The $Q$-ary code symbols are mapped to a $Q$-ary orthogonal signal set and sent over an AWGN channel with noncoherent reception. We focus on $Q = 4$, rate-1/2 convolutional codes in our designs. The random coding union bound and normal approximation are used in earlier works as benchmarks for performance for distance-spectrum optimal convolutional codes. We derive a saddlepoint approximation of the random coding union bound for the coded noncoherent signaling channel, as well as a normal approximation for this channel, and compare the performance of our codes to these limits. Our best design is within $0.6$ dB of the RCU bound at a frame error rate of $10^{-4}$.