Search

Machine learning models underpin many modern financial systems for use cases such as fraud detection and churn prediction. Most are based on supervised learning with hand-engineered features, which relies heavily on the availability of labelled data. Large self-supervised generative models have shown tremendous success in natural language processing and computer vision, yet so far they haven't been adapted to multivariate time series of financial transactions. In this paper, we present a generative pretraining method that can be used to obtain contextualised embeddings of financial transactions. Benchmarks on public datasets demonstrate that it outperforms state-of-the-art self-supervised methods on a range of downstream tasks. We additionally perform large-scale pretraining of an embedding model using a corpus of data from 180 issuing banks containing 5.1 billion transactions and apply it to the card fraud detection problem on hold-out datasets. The embedding model significantly improves value detection rate at high precision thresholds and transfers well to out-of-domain distributions.

Global financial crime activity is driving demand for machine learning solutions in fraud prevention. However, prevention systems are commonly serviced to financial institutions in isolation, and few provisions exist for data sharing due to fears of unintentional leaks and adversarial attacks. Collaborative learning advances in finance are rare, and it is hard to find real-world insights derived from privacy-preserving data processing systems. In this paper, we present a collaborative deep learning framework for fraud prevention, designed from a privacy standpoint, and awarded at the recent PETs Prize Challenges. We leverage latent embedded representations of varied-length transaction sequences, along with local differential privacy, in order to construct a data release mechanism which can securely inform externally hosted fraud and anomaly detection models. We assess our contribution on two distributed data sets donated by large payment networks, and demonstrate robustness to popular inference-time attacks, along with utility-privacy trade-offs analogous to published work in alternative application domains.

Predictive uncertainties in classification tasks are often a consequence of model inadequacy or insufficient training data. In popular applications, such as image processing, we are often required to scrutinise these uncertainties by meaningfully attributing them to input features. This helps to improve interpretability assessments. However, there exist few effective frameworks for this purpose. Vanilla forms of popular methods for the provision of saliency masks, such as SHAP or integrated gradients, adapt poorly to target measures of uncertainty. Thus, state-of-the-art tools instead proceed by creating counterfactual or adversarial feature vectors, and assign attributions by direct comparison to original images. In this paper, we present a novel framework that combines path integrals, counterfactual explanations and generative models, in order to procure attributions that contain few observable artefacts or noise. We evidence that this outperforms existing alternatives through quantitative evaluations with popular benchmarking methods and data sets of varying complexity.

Carbon-based semiconductors such as conjugated organic polymers are of potential use in the development of spintronic devices and spin-based information processing. In particular, these materials offer a low spin–orbit coupling strength due to their relatively light constituent chemical elements, which should, in principle, favour long spin diffusion lengths. However, organic polymers are relatively disordered materials and typically have a carrier mobility that is orders of magnitude lower than crystalline inorganic materials. As a result, small spin diffusion lengths of around 50 nm have typically been measured using vertical organic spin valves. Here, we report measuring spin diffusion lengths in doped conjugated polymers using a lateral spin transport device architecture, which is based on spin pumping injection and inverse spin Hall detection. The approach suggests that long spin diffusion lengths of more than 1 μm and fast spin transit times of around 10 ns are possible in conjugated polymer systems when they have a sufficiently high spin density (around 1020 cm−3). We explain these results in terms of an exchange-based spin diffusion regime in which the exchange interactions decouple spin and charge transport., info:eu-repo/semantics/published

An original concept of Slimfloor structure made of hollow-core slabs with structural concrete topping and post-tensioned delta-shaped beams is given in the paper. The technological assumptions and the results of computational analysis for this solution are also discussed. The multi-span continuous prestressed beam can significantly increase the length of the span of the Slimfloor ceiling and at the same time keeps its primary asset - almost flat bottom surface - creating the ideal solution for the demanding projects.

A short introduction to the reading seminar on dentology (the ontology of den, of “less than nothing”)., A short introduction to the reading seminar on dentology (the ontology of den, of “less than nothing”).

Tekst referuje interpretację demokrytejskiego atomizmu, przedstawioną w książce francusko-niemieckiego filozofa Heinza Wismanna, Les Avatars du vide. Interpretacja ta odegrała w ostatnich latach ważną rolę dla filozofii lacanowskiej, odkąd posłużyła Barbarze Cassin do obrony jej tezy o radykalnie nie-arystotelejskim charakterze teorii Lacana. W ślad za Cassin podążyli przedstawiciele „grupy słoweńskiej”, wykorzystując nowe odczytanie Demokryta do modyfikacji własnej teorii., This article enters into dialog with Heinz Wismann’s interpretation of the Democritean atomism presented in the book Les Avatars du Vide. Ever since Barbara Cassin referred to Wismann’s work to defend her thesis of the radically non-Aristotelian character of Lacan’s thought, Wismann’s interpretation began to play a vital role in understanding Lacanian philosophy. Cassin has been followed by members of the “Slovenian group,” who used a new reading of Democritus to modify their theory.

A short introduction to the reading seminar on dentology (the ontology of den, of “less than nothing”)., A short introduction to the reading seminar on dentology (the ontology of den, of “less than nothing”).

This article enters into dialog with Heinz Wismann’s interpretation of the Democritean atomism presented in the book Les Avatars du Vide. Ever since Barbara Cassin referred to Wismann’s work to defend her thesis of the radically non-Aristotelian character of Lacan’s thought, Wismann’s interpretation began to play a vital role in understanding Lacanian philosophy. Cassin has been followed by members of the “Slovenian group,” who used a new reading of Democritus to modify their theory., Tekst referuje interpretację demokrytejskiego atomizmu, przedstawioną w książce francusko-niemieckiego filozofa Heinza Wismanna, Les Avatars du vide. Interpretacja ta odegrała w ostatnich latach ważną rolę dla filozofii lacanowskiej, odkąd posłużyła Barbarze Cassin do obrony jej tezy o radykalnie nie-arystotelejskim charakterze teorii Lacana. W ślad za Cassin podążyli przedstawiciele „grupy słoweńskiej”, wykorzystując nowe odczytanie Demokryta do modyfikacji własnej teorii.