Search

Complexity science, despite its broad scope and potential impact, has not kept pace with fields like artificial intelligence, biotechnology and social sciences in addressing ethical concerns. The field lacks a comprehensive ethical framework, leaving us, as a community, vulnerable to ethical challenges and dilemmas. Other areas have gone through similar experiences and created, with discussions and working groups, their guides, policies and recommendations. Therefore, here we highlight the critical absence of formal guidelines, dedicated ethical committees, and widespread discussions on ethics within the complexity science community. Drawing on insights from the disciplines mentioned earlier, we propose a roadmap to enhance ethical awareness and action. Our recommendations include (i) initiating supportive mechanisms to develop ethical guidelines specific to complex systems research, (ii) creating open-access resources, and (iii) fostering inclusive dialogues to ensure that complexity science can responsibly tackle societal challenges and achieve a more inclusive environment. By initiating this dialogue, we aim to encourage a necessary shift in how ethics is integrated into complexity research, positioning the field to address contemporary challenges more effectively., Comment: 7 pages, 2 figures, 1 Annexus, 1 table

Exchange bias is applied ubiquitously throughout the spintronics industry for its ability to provide a reference direction robust to external magnetic field disturbances. For some applications, reorienting the exchange-biased reference axis may add to or enhance functionality of devices. Conventional methods for achieving this reorientation require assisting magnetic fields with long setting times and are thus limited in terms of speed and versatility. In this study, we show that by integrating functionalized materials (Co/Gd multilayers) into an existing thin film stack (Pt/Co/IrMn), perpendicular exchange bias can be reoriented $180^\circ$ by exciting it with a single femtosecond laser pulse, without an external field and on an ultrafast timescale. We found this effect to be repeatable arbitrarily many times, with only the first laser pulse incurring a loss of the initial exchange bias magnitude, yet preserving the unidirectionality required for referencing. By combining an experimental mapping of three key parameters (stack, laser and magnetic field properties) with microscopic models we were able to unravel the processes and features of the exchange bias reversal by considering the relevance of Curie, N\'eel and blocking temperatures. The outcomes of this study provide guiding principles towards designing fully optically reprogrammable magnetic sensing devices., Comment: 15 pages, 11 figures. To be submitted

For the calibration of the parameters in static and dynamic SABR stochastic volatility models, we propose the application of the GPU technology to the Simulated Annealing global optimization algorithm and to the Monte Carlo simulation. This calibration has been performed for EURO STOXX 50 index and EUR/USD exchange rate with an asymptotic formula for volatility or Monte Carlo simulation. Moreover, in the dynamic model we propose an original more general expression for the functional parameters, specially well suited for the EUR/USD exchange rate case. Numerical results illustrate the expected behavior of both SABR models and the accuracy of the calibration. In terms of computational time, when the asymptotic formula for volatility is used the speedup with respect to CPU computation is around $200$ with one GPU. Furthermore, GPU technology allows the use of Monte Carlo simulation for calibration purposes, the computational time with CPU being prohibitive.

The present work addresses the challenge of training neural networks for Dynamic Initial Margin (DIM) computation in counterparty credit risk, a task traditionally burdened by the high costs associated with generating training datasets through nested Monte Carlo (MC) simulations. By condensing the initial market state variables into an input vector, determined through an interest rate model and a parsimonious parameterization of the current interest rate term structure, we construct a training dataset where labels are noisy but unbiased DIM samples derived from single MC paths. A multi-output neural network structure is employed to handle DIM as a time-dependent function, facilitating training across a mesh of monitoring times. The methodology offers significant advantages: it reduces the dataset generation cost to a single MC execution and parameterizes the neural network by initial market state variables, obviating the need for repeated training. Experimental results demonstrate the approach's convergence properties and robustness across different interest rate models (Vasicek and Hull-White) and portfolio complexities, validating its general applicability and efficiency in more realistic scenarios.

In these notes we propose and analyze an inertial type method for obtaining stable approximate solutions to nonlinear ill-posed operator equations. The method is based on the Levenberg-Marquardt (LM) iteration. The main obtained results are: monotonicity and convergence for exact data, stability and semi-convergence for noisy data. Regarding numerical experiments we consider: i) a parameter identification problem in elliptic PDEs, ii) a parameter identification problem in machine learning; the computational efficiency of the proposed method is compared with canonical implementations of the LM method.

This paper addresses the problem of pricing involved financial derivatives by means of advanced of deep learning techniques. More precisely, we smartly combine several sophisticated neural network-based concepts like differential machine learning, Monte Carlo simulation-like training samples and joint learning to come up with an efficient numerical solution. The application of the latter development represents a novelty in the context of computational finance. We also propose a novel design of interdependent neural networks to price early-exercise products, in this case, Bermudan swaptions. The improvements in efficiency and accuracy provided by the here proposed approach is widely illustrated throughout a range of numerical experiments. Moreover, this novel methodology can be extended to the pricing of other financial derivatives.

MRI systems are traditionally engineered to produce close to idealized performance, enabling a simplified pulse sequence design philosophy. An example of this is control of eddy currents produced by gradient fields; usually these are compensated by pre-emphasizing demanded waveforms. This process typically happens invisibly to the pulse designer, allowing them to assume the achieved gradient waveform will be as desired. Whilst convenient, this imposes stricter limits on the sequence design than the hardware can handle (for example, pre-emphasis adds an additional overhead to amplifiers). This strategy can be undesirable particularly for lower performance or resource-limited hardware. Instead we explore the use of a 'digital twin' (i.e. an end-to-end model of the scanner system) to optimize control inputs, resulting in sequences that inherently compensate for known imperfections. We explore digital twin optimization specifically for gradient system imperfections as an exemplar. This is first explored in simulations using a simple exponential eddy current model, then experimentally using an empirical gradient impulse response function on a 7T MRI system. When unconstrained, digital twin optimization reproduces classic pre-emphasis. When strict hardware constraints are imposed (simulating lower performance hardware), it identifies novel sequences for scenarios where classic pre-emphasis would be unachievable. Experimentally, the optimization approach was demonstrated to substantially reduce ghosting effects in echo planar images on a 7T system. Digital twin optimization may allow more efficient use of hardware by taking a whole system approach. Ultimately this could enable the use of cheaper hardware without loss of performance., Comment: 33 pages, 14 figures (including supporting information)

Learning to defer (L2D) aims to improve human-AI collaboration systems by learning how to defer decisions to humans when they are more likely to be correct than an ML classifier. Existing research in L2D overlooks key real-world aspects that impede its practical adoption, namely: i) neglecting cost-sensitive scenarios, where type I and type II errors have different costs; ii) requiring concurrent human predictions for every instance of the training dataset; and iii) not dealing with human work-capacity constraints. To address these issues, we propose the \textit{deferral under cost and capacity constraints framework} (DeCCaF). DeCCaF is a novel L2D approach, employing supervised learning to model the probability of human error under less restrictive data requirements (only one expert prediction per instance) and using constraint programming to globally minimize the error cost, subject to workload limitations. We test DeCCaF in a series of cost-sensitive fraud detection scenarios with different teams of 9 synthetic fraud analysts, with individual work-capacity constraints. The results demonstrate that our approach performs significantly better than the baselines in a wide array of scenarios, achieving an average $8.4\%$ reduction in the misclassification cost. The code used for the experiments is available at https://github.com/feedzai/deccaf

In this manuscript we propose and analyze an implicit two-point type method (or inertial method) for obtaining stable approximate solutions to linear ill-posed operator equations. The method is based on the iterated Tikhonov (iT) scheme. We establish convergence for exact data, and stability and semi-convergence for noisy data. Regarding numerical experiments we consider: i) a 2D Inverse Potential Problem, ii) an Image Deblurring Problem; the computational efficiency of the method is compared with standard implementations of the iT method., Comment: 22 pages, 8 figures. Accepted for publication at Inverse Problems

Public dataset limitations have significantly hindered the development and benchmarking of learning to defer (L2D) algorithms, which aim to optimally combine human and AI capabilities in hybrid decision-making systems. In such systems, human availability and domain-specific concerns introduce difficulties, while obtaining human predictions for training and evaluation is costly. Financial fraud detection is a high-stakes setting where algorithms and human experts often work in tandem; however, there are no publicly available datasets for L2D concerning this important application of human-AI teaming. To fill this gap in L2D research, we introduce the Financial Fraud Alert Review Dataset (FiFAR), a synthetic bank account fraud detection dataset, containing the predictions of a team of 50 highly complex and varied synthetic fraud analysts, with varied bias and feature dependence. We also provide a realistic definition of human work capacity constraints, an aspect of L2D systems that is often overlooked, allowing for extensive testing of assignment systems under real-world conditions. We use our dataset to develop a capacity-aware L2D method and rejection learning approach under realistic data availability conditions, and benchmark these baselines under an array of 300 distinct testing scenarios. We believe that this dataset will serve as a pivotal instrument in facilitating a systematic, rigorous, reproducible, and transparent evaluation and comparison of L2D methods, thereby fostering the development of more synergistic human-AI collaboration in decision-making systems. The public dataset and detailed synthetic expert information are available at: https://github.com/feedzai/fifar-dataset, Comment: The public dataset and detailed synthetic expert information are available at: https://github.com/feedzai/fifar-dataset

We provide quantitative evidence suggesting social learning in sperm whales across socio-cultural boundaries, using acoustic data from the Pacific and Atlantic Oceans. Traditionally, sperm whale populations are categorized into clans based on their vocal repertoire: the rhythmically patterned click sequences (codas) that they use. Among these codas, identity codas function as symbolic markers for each clan, accounting for 35-60% of codas they produce. We introduce a computational method to model whale speech, which encodes rhythmic micro-variations within codas, capturing their vocal style. We find that vocal style-clans closely align with repertoire-clans. However, contrary to vocal repertoire, we show that sympatry increases vocal style similarity between clans for non-identity codas, i.e. most codas, suggesting social learning across cultural boundaries. More broadly, this subcoda structure model offers a framework for comparing communication systems in other species, with potential implications for deeper understanding of vocal and cultural transmission within animal societies.

The use of machine learning for time series prediction has become increasingly popular across various industries thanks to the availability of time series data and advancements in machine learning algorithms. However, traditional methods for time series forecasting rely on pre-optimized models that are ill-equipped to handle unpredictable patterns in data. In this paper, we present EAMDrift, a novel method that combines forecasts from multiple individual predictors by weighting each prediction according to a performance metric. EAMDrift is designed to automatically adapt to out-of-distribution patterns in data and identify the most appropriate models to use at each moment through interpretable mechanisms, which include an automatic retraining process. Specifically, we encode different concepts with different models, each functioning as an observer of specific behaviors. The activation of the overall model then identifies which subset of the concept observers is identifying concepts in the data. This activation is interpretable and based on learned rules, allowing to study of input variables relations. Our study on real-world datasets shows that EAMDrift outperforms individual baseline models by 20% and achieves comparable accuracy results to non-interpretable ensemble models. These findings demonstrate the efficacy of EAMDrift for time-series prediction and highlight the importance of interpretability in machine learning models., Comment: Submitted to ECML PKDD 2023