Generic AI Research Articles

Communicating Banking Cyber-security Measures, Customer Ethical Concerns, Experience, and Loyalty Intentions: A Developing Economy’s Perspective

AI technology-based banking services development has disrupted the way people participate in banking transactions. It has created easier and faster banking transaction possibilities with the use of electronic gadgets. However, ethical concerns about these applications have also been amplified together with the need for management communication of safety features and protocols for customer information protection, and redress when infringements occur. The study was an attempt to highlight how AI-enabled banking services safety communication affects customers’ ethical concerns and how the concerns shape their banking services value perception, attitude, and loyalty intentions. A conceptual framework based on the generic AI technology, ethical concerns, and loyalty intentions was used as a basis for this study. It attempted to test the link between management communication, ethical concerns, satisfaction/dissatisfaction, and customer loyalty to AI-based banking services in a developing economy context. The study used three theoretical grounding bases to empirically test the proposed hypotheses. The results analysis followed Structural equation modeling (SEM). The results confirmed the impact of management communication on customers’ ethical concerns of security, privacy, diversity, and discrimination, and the positive influence of privacy and security on satisfaction/dissatisfaction. However, the relationship between diversity and discrimination concerns with customer satisfaction was not confirmed. Lastly, customer satisfaction was proven to impact their loyalty intentions.

Developing a holistic AI literacy assessment matrix – Bridging generic, domain-specific, and ethical competencies

Motivated by a holistic understanding of AI literacy, this work presents an interdisciplinary effort to make AI literacy measurable in a comprehensive way, considering generic and domain-specific AI literacy as well as AI ethics. While many AI literacy assessment tools have been developed in the last 2-3 years, mostly in the form of self-assessment scales and less frequently as knowledge-based assessments, previous approaches only accounted for one specific area of a comprehensive understanding of AI competence, namely cognitive aspects within generic AI literacy. Considering the demand for AI literacy development for different professional domains and reflecting on the concept of competence in a way that goes beyond mere cognitive aspects of conceptual knowledge, there is an urgent need for assessment methods that capture domain-specific AI literacy on each of the three competence dimensions of cognition, behavior, and attitude. In addition, competencies for AI ethics are becoming more apparent, which further calls for a comprehensive assessment of AI literacy for this very matter. This conceptual paper aims to provide a foundation upon which future AI literacy assessment instruments can be built and provides insights into what a framework for item development might look like that addresses both generic and domain-specific aspects of AI literacy as well as AI ethics literacy, and measures more than just knowledge-related aspects based on a holistic approach.

Operational AI analysis flows: A framework for the efficient integration and execution within test processes

The European Commission’s “Flightpath 2050” highlights the critical role of the design phase in complex aircraft development. Ensuring the aircraft’s functionality, particularly for certification, represents a significant proportion (around 50%) of the total development cost. The increasing complexity of the testing process presents challenges. For example, the creation of simulation models for HIL testing and the performance of exploratory tests for certification are becoming increasingly complex due to the growing complexity of aircraft design.To address this challenge, this research aims to integrate AI modelling and data analysis into aircraft test processes as part of the design process. The aim is to develop a user-friendly modelling tool for test engineers without the need for data analyst expertise. The approach focuses on achieving seamless interoperability and easy integration of test-related data sources and AI models, eliminating the need for complex scripting in scripting languages such as Python or Matlab. To achieve this goal, a domain ontology and generic wrappers have been developed to map generic AI models to the TensorFlow and Matlab tools. At the same time, data sources are connected via a data integration tool.The evaluation successfully demonstrates the ability to analyse the correlation between sensors in an experimental cooling system without understanding the data format of the data sources and without the scripting of the AI model to model and execute the AI models.

Zero-shot transfer learned generic AI models for prediction of optimally ripe climacteric fruits

Ideally, ripe fruits offer appropriate nutritional content and best quality in terms of taste and flavour. Prediction of ripe climacteric fruits acts as the main marketing indicator for quality from the consumer perspective and thus renders it a genuine industrial concern for all the stakeholders of the fruit supply chain. However, the building of fruit-specific individual model for the prediction of ripeness level remains an existing challenge due to the scarcity of sufficient labeled experimental data for each fruit. This paper describes the development of generic AI models based on the similarity in physico-chemical degradation phenomena of climacteric fruits for prediction of ‘unripe’ and ‘ripe’ levels using ‘zero-shot’ transfer learning techniques. Experiments were performed on a variety of climacteric and non-climacteric fruits, and it was observed that transfer learning works better for fruits within a cluster (climacteric fruits) as compared to across clusters (climacteric to non-climacteric fruits). The main contributions of this work are two-fold (i) Using domain knowledge of food chemistry to label the data in terms of age of the fruit, (ii) We hypothesize and prove that the zero-shot transfer learning works better within a set of fruits, sharing similar degradation chemistry depicted by their visual properties like black spot formations, wrinkles, discoloration, etc. The best models trained on banana, papaya and mango dataset resulted in s zero-shot transfer learned accuracies in the range of 70 to 82 for unknown climacteric fruits. To the best of our knowledge, this is the first study to demonstrate the same.

A generic AI pipeline for identifying known species and discovering new species

To counterattack the current biodiversity crisis, we need innovative tools and technologies that can greatly improve species identification and discovery at scale. Here, we describe a data analysis pipeline to automatically identify biological sound categories in soundscape data and accelerate the discovery of new species. The pipeline consists of two components: (1) Unsupervised audio event detection and clustering that enables automatic detection and categorization of biological sounds in large audio datasets and (2) supervised deep learning-based signal detection that enables automated cluster labeling and detection of known sound categories in new audio. We evaluated the results of our unsupervised analysis tools applied to a labeled dataset of ∼18 h of recordings from Panama. Experts assigned a label to each sound event with a max frequency of 6 kHz. Our results show that clusters generally had high homogeneity and completeness and indicate that our cluster analysis tool can successfully detect and group like-sounds together, improving data processing and species identification. In addition, we were able to implement these analytical tools in a user-friendly data analysis platform that will provide scientists, wildlife managers, conservationists, and public and private environmental organizations with the information they need to make informed conservation and management decisions.

Medication delivery factors and adjuvant endocrine therapy adherence in breast cancer.

Over 50% of breast cancer patients prescribed a 5-year course of daily oral adjuvant endocrine therapy (ET) are nonadherent. We investigated the role of costs and cancer medication delivery mode and other medication delivery factors on adherence. We conducted a retrospective cohort study of commercially insured and Medicare advantage patients with newly diagnosed breast cancer in 2007-2015 who initiated ET. We examined the association between 12-month ET adherence (proportion of days covered by fills ≥ 0.80) and ET copayments, 90-day prescription refill use, mail order pharmacy use, number of pharmacies, and synchronization of medications. We used regression models to estimate nonadherence risk ratios adjusted for demographics (age, income, race, urbanicity), comorbidities, total medications, primary cancer treatments, and generic AI availability. Sensitivity analyses were conducted using alternative specifications for independent variables. Mail order users had higher adherence in both commercial and Medicare-insured cohorts. Commercially insured patients who used mail order were more likely to be adherent if they had low copayments (< $5) and 90-day prescription refills. For commercially insured patients who used local pharmacies, use of one pharmacy and better synchronized refills were also associated with adherence. Among Medicare patients who used mail order pharmacies, only low copayments were associated with adherence, while among Medicare patients using local pharmacies both low copayments and 90-day prescriptions were associated with ET adherence. Out-of-pocket costs, medication delivery mode, and other pharmacy-related medication delivery factors are associated with adherence to breast cancer ET. Future work should investigate whether interventions aimed at streamlining medication delivery could improve adherence for breast cancer patients.

Generic AI models for mass transfer coefficient prediction in amine-based CO2 absorber, Part I: BPNN model

Accurate and reliable prediction of mass transfer coefficient is critical to evaluate the mass transfer performance of amine-based carbon capture process. This work aims to establish a generic mass transfer model applicable to CO2 absorption into amine solutions in absorber columns. A series of back-propagation neural network (BPNN) models were established based on the experimental data of 23 amine-based systems. The BPNN models trained for a specific CO2 absorption process matched well with the experimental data with all the AAREs of below 4%. Meanwhile, models with better applicability were proposed and established by introducing extra parameters related to amine properties or column characteristics to better describe and predict the mass transfer behavior in different amine-based systems. Particularly, the generic mass transfer coefficient model was developed with the consideration of all influencing factors including operating conditions, amine properties, and packing characteristics, to achieve the accurate prediction of mass transfer coefficient of CO2 absorption into amine solutions in packed columns. The developed generic model could serve as a fast and efficient tool for preliminary selection and evaluation of potential amines by presenting the mass transfer performance in CO2 absorber.

Generic AI models for mass transfer coefficient prediction in amine‐based CO2 absorber, Part II: RBFNN and RF model

AbstractIn this work, the radial basis function neural network (RBFNN) and random forest (RF) algorithms were employed to develop generic AI models predicting mass transfer coefficient in amine‐based CO2 absorber. The models with operating parameters as input gave quite different prediction performance in different CO2 absorption systems. To secure better applicability, extra parameters related to amine type and packing characteristics were introduced to reasonably describe mass transfer behaviors, respectively. Moreover, the generic models were proposed by considering all influencing factors of mass transfer in CO2 absorber column. Furthermore, the performance of BPNN, RBFNN, and RF models was completely compared and fully discussed in terms of AARE. All three generic models could predict mass transfer coefficient of CO2 absorber very well. It was found that the BPNN models provide the best predication with AAREs of below 5%. The developed generic model could serve as a fast and efficient tool for preliminary selection and evaluation of potential amines for CO2 absorption. The framework of generic ML model development was also clearly presented, which could provide theoretical basis and practical guidance for the implementation and application of ML models in the carbon capture field.

Factors associated with adjuvant endocrine therapy adherence in non-metastatic breast cancer.

533 Background: Over 50% of breast cancer patients prescribed adjuvant endocrine therapy (ET) are nonadherent during the recommended 5-year course of therapy. We investigated the role of cancer medication delivery on adherence, including mail order pharmacy use, number of pharmacies and copays. Methods: We conducted a retrospective cohort study of 15,769 commercially insured breast cancer patients age 18-64 with newly diagnosed breast cancer in 2007-2015 that initiated ET. Incident breast cancer was identified by a validated algorithm which identifies mainly non-metastatic breast cancer. We examined the association between 12-month AET adherence (proportion of days covered by fills ≥0.80) and mail order pharmacy use, number of pharmacies, and AET copays. We used Poisson regression to estimate nonadherence risk ratios and adjusted for demographics (age, income, race, urbanicity), comorbidities, total medications, primary cancer treatments (surgery, radiation, chemo, and ET initiated), and generic AI availability. Sensitivity analyses were conducted using alternate specifications for independent variables. To test whether any observed differences were due to self-selection, we also conducted a negative control analysis. Results: Most patients were white (74.4%) and age 55-64 (43.3%). Only 16% of patients used a mail order pharmacy for ET fills, most patients only used one pharmacy (58.8%) and 25.2% had a co-pay of $20 or more. In the primary analysis, mail order patients were more likely to be adherent to their ET (aRR 1.21; 95% CI 1.18-1.24), patients using one pharmacy were more likely to be adherent (1 vs 3+: aRR 1.09; 95% CI 1.06-1.13), and patients with lower copays were more likely to be adherent (quartile 1 vs 4: aRR 1.04; 95% CI 1.01-1.08). Results were consistent across sensitivity analyses, and there was no association between mail order and copays and the negative control outcome of any pneumonia diagnoses. Conclusions: Medication delivery factors are associated with adherence to breast cancer AET. Future work should investigate whether interventions to streamline medication delivery could improve adherence for this population.

Actionable Ethics through Neural Learning

While AI is going to produce a great impact on society, its alignment with human values and expectations is an essential step towards a correct harnessing of AI potentials for good. There is a corresponding growing need for mature and established technical standards to enable the assessment of an AI application as the evaluation of its graded adherence to formalized ethics. This is clearly dependent on methods to inject ethical awareness at all stages of an AI application development and use. For this reason we introduce the notion of Embedding Principles of ethics by Design (EPbD) as a comprehensive inductive framework. Although extending generic AI applications, it mainly aims at learning the ethical behaviour through numerical optimization, i.e. deep neural models. The core idea is to support ethics by integrating automated reasoning over formal knowledge and induction from ethically enriched training data. A deep neural network is proposed here to model both the functional as well as the ethical conditions characterizing a target decision. In this way, the discovery of latent ethical knowledge is enabled and made available to the learning process. The application of the above framework to a banking application, i.e. AI-driven Digital Lending, is used to show how accurate classification can be achieved without neglecting the ethical dimension. Results over existing datasets demonstrate that the ethical compliance of the sources can be used to output models able to optimally fine tune the balance between business and ethical accuracy.

Enabling Serverless Deployment of Large-Scale AI Workloads

We propose a set of optimization techniques for transforming a generic AI codebase so that it can be successfully deployed to a restricted serverless environment, without compromising capability or performance. These involve (1) slimming the libraries and frameworks (e.g., pytorch) used, down to pieces pertaining to the solution; (2) dynamically loading pre-trained AI/ML models into local temporary storage, during serverless function invocation; (3) using separate frameworks for training and inference, with ONNX model formatting; and, (4) performance-oriented tuning for data storage and lookup. The techniques are illustrated via worked examples that have been deployed live on geospatial data from the transportation domain. This draws upon a real-world case study in intelligent transportation looking at on-demand, realtime predictions of flows of train movements across the UK rail network. Evaluation of the proposed techniques shows the response time, for varying volumes of queries involving prediction, to remain almost constant (at 50 ms), even as the database scales up to the 250M entries. The query response time is important in this context as the target is predicting train delays. It is even more important in a serverless environment due to the stringent constraints on serverless functions’ runtime before timeout. The similarities of a serverless environment to other resource constrained environments (e.g., IoT, telecoms) means the techniques can be applied to a range of use cases.

Patient costs of breast cancer endocrine therapy agents under Medicare Part D vs with generic formulations.

PurposeThe high expense of newer, more effective adjuvant endocrine therapy agents (aromatase inhibitors [AIs]) for postmenopausal breast cancer contributes to socioeconomic disparities in breast cancer outcomes. This study compares endocrine therapy costs for breast cancer patients during the first five years of Medicare Part D implementation, and when generic alternatives became available.MethodsThe out of pocket patient costs for AIs and tamoxifen under Medicare Part D drug plans were determined for 2006–2011 from the CMS Website for the 50 US states and District of Columbia.ResultsBetween 2006 and 2010, the mean annual patient drug cost under Medicare Part D in the median state rose 19% for tamoxifen, 113% for anastrozole, 89% for exemestane, and 129% for letrozole, resulting in median annual out of pocket costs in 2010 of $701, $3050, $2804, and $3664 respectively. However, the 2011 availability of generic AI preparations led to median annual costs in 2011 of $804, $872, $1837, and $2217 respectively. Not included in the reported patient costs, the mean monthly drug premiums in the median state increased 58% in 2011 compared to 2007.ConclusionsThe more effective AI agents became considerably more expensive during the first several years of the Medicare Part D program. Cost decreased with the introduction of generic agents, an intervention that was independent of the Part D program. It is unlikely that the Part D program ameliorated existing socioeconomic disparities in survival among breast cancer patients, but the availability of generic agents may do so.

The Axiom General Purpose Game Playing System

A general game playing system, Axiom, was developed to overcome the limitations inherent in an existing commercial general purpose game playing system which incorporates a generic AI. Axiom extends the range of possible games via an extensible language, thereby allowing for a high degree of game-specific customization within the AI as well as in defining complex game behaviors. This paper provides an overview of the Axiom system and discusses some important aspects of the customizable AI. It also answers the question “Can a powerful game definition language be developed to enhance a popular general game platform?”

The ARPA-Rome Knowledge-Based Planning and Scheduling Initiative

In 1990, ARPA and the US Air Force Rome Laboratory formed the Knowledge-Based Planning and Scheduling Initiative. Better known as the ARPA/Rome Planning Initiative or ARPI, the program was formed to develop the next generation of generic AI tools for planning, scheduling, and resource allocation, as well as to significantly improve crisis management planning. At ARPI's core is a vision of a distributed network of graphics-based planning cells, sharing a common reasoning infrastructure that enables continuous concurrent planning. It takes a distributed, collaborative approach to crisis management, and applies total quality management techniques to ensure measurable progress and remote cooperation among participants.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Sharing the solution: The need for generic artificial intelligence decision support development tools in battle management

Abstract The Navy conducted a multi-laboratory study to investigate ways in which machine-based technology can provide intelligent decision support to Navy battle management. A key result from this study is that a need exists to develop generic AI decision support development tools for battle management. This paper describes the major requirement areas identified in the study for building such tools.