The article discusses the use of #deeplearning and #datamining in business intelligence protocols to optimize data-driven decision-making and improve efficiency. The authors focus on the use of Graph Neural Network and Autoencoders Models to process large amounts of data and model #fraud behaviors. They suggest that deep learning can be used to control #moneylaundering in financial institutions and improve visibility and transparency in businesses.
#machinelearning #algorithms are increasingly for #riskassessment in the #insuranceindustry, with hybrid methods often outperforming individual ones. Research has identified challenges such as tackling imbalanced datasets, selecting features, and improving interpretability. Newer methods such as #deeplearning and ensembles may further improve accuracy.
This article discusses the need for #regulation of #robots and #ai in #europe, focusing on the issue of #civil #liability. Despite multiple attempts to harmonize #eu#tort #law, only the liability of producers for defective products has been successfully harmonized so far. The #aiact, published by the #europeancommission in 2021, aims to #regulate AI at the European level by classifying #smartrobots as "high risk systems", but does not address liability rules. This article explores liability issues related to AI and robots, particularly when using #deeplearning #machinelearning techniques that challenge the traditional liability paradigm.
"This study proposes a comprehensive method (with representative AI-Technologies as a data basis) for the structured and targeted categorization and classification of AI under the risk-based audit approach. Initial feedback received by AI-Experts regarding the design and development of the artifact is collected. With the developed method, the study contributes to the descriptive and prescriptive knowledge base regarding the categorization and classification of AI within the auditing and accounting profession."
"We here propose a novel XAI [eXplainable AI] technique for deep learning methods (DL) which preserves and exploits the natural time ordering of the data. Simple applications to financial data illustrate the potential of the new approach in the context of risk-management and fraud-detection."
"The empirical results in this research show that the classification performance of our proposed methodology is superior compared to that of a large number of traditional classifier models. We also show that our proposed methodology solves the limitation of previous bankruptcy models using textual data, as they can only make predictions for a small proportion of companies."
"Despite that some attention or self-attention based models with time-aware or feature-aware enhanced strategies have achieved better performance compared with other temporal modeling methods, such improvement is limited due to a lack of guidance from global view. To address this issue, we propose a novel end-to-end Hierarchical Global View-guided (HGV) sequence representation learning framework. "
"... DL [deep learning] does not outperform traditional ML [machine learning] models in the case of structured datasets with fixed-length feature vectors. Deep learning should be regarded as a powerful addition to the existing body of ML models instead of a one size fits all solution."
"The role of regulatory agencies will be crucial to protect consumers while allowing innovation. There is currently no unified regulatory framework."