Explore the corpus

Find evidence, not just documents.

Search Minerva’s current Findings using transparent hybrid retrieval. Results remain strictly inside the configured corpus.

Hybrid retrieval

Results for “Miscellaneous”

30 findings

62match

Finding 23931 source

The chapter integrates deep learning with explainable AI to demystify models, combining text, audio, video, and numerical data for insights into investor sentiment, financial distress, portfolio construction, and corporate governance.

LexicalMeaningQualityGraph

Matched: miscellaneous

62match

Finding 23921 source

The study demonstrates how artificial intelligence transforms corporate finance by enhancing prediction accuracy, uncovering hidden patterns, and analyzing unstructured data with models like random forests, neural networks, and causal forests.

LexicalMeaningQualityGraph

Matched: miscellaneous

61match

Finding 78411 source

Researchers introduce a novel financial corpus and a glossary, paired with a methodology to analyze term match accuracy at corpus and segment levels, outperforming large language models on financial content.

LexicalMeaningQualityGraph

Matched: miscellaneous

61match

Finding 23941 source

Employing ensemble learning, SHAP values, and double machine learning, the research boosts performance while noting challenges like data manipulation, fake citations, and quality compromises, urging human-machine collaboration and market impact.

LexicalMeaningQualityGraph

Matched: miscellaneous

61match

Finding 78421 source

Combining advanced automatic alignment with manually curated glossaries, the study provides statistical analyses revealing that morphological complexity significantly influences translation metrics, while acknowledging limitations focused on macroeconomic texts in practice.

LexicalMeaningQualityGraph

Matched: miscellaneous

60match

Finding 44701 source

The study reveals that traditional fraud detection methods falter against sophisticated schemes, while the novel hybrid framework enhances accuracy, adaptability, and scalability by leveraging relational data analysis and reinforcement techniques.

LexicalMeaningQualityGraph

Matched: miscellaneous

60match

Finding 78401 source

The study reveals that domain-specific financial terminology critically impacts machine translation quality across 22 European languages, showing surprising strong correlations between terminology accuracy and overall performance in specialized financial texts.

LexicalMeaningQualityGraph

Matched: miscellaneous

60match

Finding 80301 source

Employing advanced techniques including the Central Limit Theorem, spectral analysis, and truncation arguments, researchers validate the method's robustness and demonstrate its real-world applicability to data such as Japans consumption figures.

LexicalMeaningQualityGraph

Matched: miscellaneous

60match

Finding 44711 source

Extensive experiments with real-world financial datasets expose surprising trends of improved detection rates and fewer false positives, despite computational overhead, establishing a valuable foundation for ongoing research in fraud prevention.

LexicalMeaningQualityGraph

Matched: miscellaneous

60match

Finding 34951 source

Key contributions include a defined protocol for reasoning generation, structured output formatting with distinct tags, innovative compliance management via multi-agent synthesization, while noting language diversity and reinforcement learning challenges effectively.

LexicalMeaningQualityGraph

Matched: miscellaneous

60match

Finding 44691 source

Researchers from Carnegie Mellon, Bentley, Cornell, UT Dallas, and Renmin unveil a cutting-edge approach integrating Graph Neural Networks with Reinforcement Learning to detect complex financial fraud in dynamic, interconnected systems.

LexicalMeaningQualityGraph

Matched: miscellaneous

60match

Finding 80311 source

Despite its technical complexity, the chapter refines econometric techniques and introduces new terminologies that clarify economic signal decomposition, overcoming non-stationary process challenges and marking significant advances in macroeconomic data analysis.

LexicalMeaningQualityGraph

Matched: miscellaneous

60match

Finding 34931 source

The paper introduces DianJin-R1, a framework embedding explicit reasoning within financial large language models, dramatically enhancing accuracy and interpretability through step-by-step logical explanations built from diverse, high-quality datasets for innovation.

LexicalMeaningQualityGraph

Matched: miscellaneous

60match

Finding 67721 source

Researchers fused reinforcement learning with behavioral economics to transform AML detection, achieving a 27.4% boost in detection accuracy and an 18.6% drop in false alerts, surpassing conventional methods across industries.

LexicalMeaningQualityGraph

Matched: miscellaneous

60match

Finding 25811 source

Innovative terminologies such as transitory efficiency and persistent efficiency enrich economic discussions on cost management, with this methodology paving the way for future exploration of model uncertainty and banking applications.

LexicalMeaningQualityGraph

Matched: miscellaneous

60match

Finding 34941 source

Surprisingly, even smaller models augmented with reasoning and a novel Group Relative Policy Optimization reinforcement learning technique outperformed larger non-reasoning models, matching multi-agent system performance while dramatically reducing computational costs.

LexicalMeaningQualityGraph

Matched: miscellaneous

60match

Finding 34211 source

DeepSeek, a Chinese AI firm, pioneered a breakthrough FinTech model that slashes development costs from $100 million to $6 million, challenging traditional GPT-4 systems and democratizing AI through open-source innovation.

LexicalMeaningQualityGraph

Matched: miscellaneous

60match

Finding 67741 source

Extensive experiments on synthetic and real-world datasets validate the innovative method; however, high computational demands and scalability challenges necessitate further research and optimization in AML practices for enhanced efficiency urgently.

LexicalMeaningQualityGraph

Matched: miscellaneous

60match

Finding 34221 source

Utilizing innovative techniques including multi-head latent attention and a mixture of experts, DeepSeek attains high performance with minimal resources, promoting sustainability and empowering startups and smaller firms with democratized technology.

LexicalMeaningQualityGraph

Matched: miscellaneous

59match

Finding 80291 source

A groundbreaking SIML filtering method is introduced that extracts hidden trend-cycle, seasonal, and measurement error components from noisy, non-stationary economic time series using frequency domain analysis and asymptotic theoretical rigor.

LexicalMeaningQualityGraph

Matched: miscellaneous

59match

Finding 34231 source

Research reveals surprising market disruption with inclusivity in lending, investment and fraud detection, yet warns of risks including censorship, security vulnerabilities, regulatory challenges and geopolitical tensions, necessitating coordinated global oversight.

LexicalMeaningQualityGraph

Matched: miscellaneous

59match

Finding 23951 source

The paper presents an original synthesis of established corporate finance methods with innovative applications of $AI$ and $machine learning$. It introduces fresh perspectives on analyzing unstructured data, causal inference, and related issues. Though primarily building on existing ideas, its thoughtful integration offers readers valuable insights and compelling practical relevance overall.

LexicalMeaningQualityGraph

Matched: miscellaneous

59match

Finding 25801 source

The paper presents state-of-the-art techniques including advanced MCMC methods and extensive simulation studies, although its reliance on large datasets and technical complexity may restrict immediate practical application for industry practitioners.

LexicalMeaningQualityGraph

Matched: miscellaneous

59match

Finding 67731 source

Authors introduce a unified framework that models AML as a sequential decision-making process, integrating composite reward functions and dual-process architectures to accurately differentiate legitimate from suspicious transactions across borders now.

LexicalMeaningQualityGraph

Matched: miscellaneous

59match

Finding 21331 source

Authors introduce novel concepts such as Autonomous Cognitive Twins and Quantum Financial Twins, shifting paradigms from static replicas to dynamic agents through deep reinforcement learning, neuromorphic computing, and federated learning.

LexicalMeaningQualityGraph

Matched: miscellaneous

59match

Finding 21321 source

Study finds Financial Digital Twins boost customer experience by 30--40% while mitigating risks, demonstrating a multi-layer architecture integrating data fabrics, behavioral models, simulations, decision intelligence, bi-directional synchronization, and predictive agency.

LexicalMeaningQualityGraph

Matched: miscellaneous

59match

Finding 21311 source

Researchers reveal that integrating artificial intelligence with digital twin technology enables real-time, customer-centric financial models, dynamically predicting behaviors and reducing operational risks, marking a breakthrough transformation in personalized financial services.

LexicalMeaningQualityGraph

Matched: miscellaneous

59match

Finding 25791 source

The study introduces a dynamic Bayesian nonparametric framework using the Logit Stick-Breaking Process, distinguishing transient and persistent cost inefficiencies and showing banks switch models during major shocks like COVID-19 recently.

LexicalMeaningQualityGraph

Matched: miscellaneous

59match

Finding 30951 source

Case studies in the financial sector demonstrate auditing evaluating AI systems on fairness, bias, explainability, accountability, and robustness, while pilot programs utilize seven-domain criteria with document reviews and stakeholder interviews.

LexicalMeaningQualityGraph

Matched: miscellaneous

58match

Finding 67221 source

A state-of-the-art Temporal Graph Neural Network model integrates spatial and temporal features to achieve fraud detection accuracy of 99.24% and 98.76%, while reducing false positives by 37% in cross-border payments.

LexicalMeaningQualityGraph

Matched: miscellaneous