This article is a contribution within the Erasmus+ project titled “The Future Lies in Applied Artificial Intelligence(FAAI) and examines research of collected IT specifications of good practices in Area of Artificial Intelligence (AAI). The article describes research conducted, the purpose of which is to find IT specifications of good practices in AI and describe their characteristics, like an area of implementation of the AI solution, the result of processing the data, the source of data, Data processing, and quality, what tools are used for processing data, and others. AAI application cases and the technologies used for implementation are reviewed. The specifics of the data and the applications used are described. The examination of these technologies will provide insight into which ones are favored and provide an overview of what is commonly referred to as “best practices” in this particular domain.The research encompassed a global examination of cases. The analysis of the data offers valuable insights in various directions:

 Application area of ML/AI

 Type of machine learning problems in described good

practices in Artificial Intelligence

 Type of models were developed within the projects

 What is the area of implementation of AI solution

 Used AI libraries (frameworks).

 Source of data

 Data characteristics

 Tools are used to store data

 What platform solution is used

 What type of storage is used.

 

The full paper can be found at: https://ieeexplore.ieee.org/document/10316145

Audio version is available:

The work is fulfilled within the framework of Erasmus+ project “The Future is in Applied Artificial Intelligence” (FAAI) and devoted to the development the methodology for collecting and analyzing good practices in the field of applied artificial intelligence (AAI) regarding the competences, training, existing solutions and real cases, which can be used for developing training courses of competence based education. Here we propose the definition of good practice in the field of AAI together with the corresponding criteria and features. The offered methodology uses system research based on the data gathered from existing training courses in AAI, labor market, surveys filled in by academics, students and employers, AAI use cases in science and industry.

The full paper can be found at: https://ieeexplore.ieee.org/document/10316104

Audio version is available:

The article is devoted to the problem of developing a mathematical model of the response of a potentiometric biosensor for the determination of α-chaconine in the form of a system of seven differential equations that describe the dynamics of biochemical reactions during the full cycle of α-chaconine concentration measurement. At the same time, each of the differential equations establishes the concentration dependence of substrate, enzyme, inhibitor, enzyme-substrate, product, enzyme-inhibitor, enzyme-substrate-inhibitor complexes as a function of time. The mathematical model of the biosensor for the determination of α-chaconine was solved numerically in the R package. The input parameters of the system were used, namely, the concentrations of the enzyme, substrate, and inhibitor (5.8×10-4 M butyrylcholinesterase, 1×10-3 M butyrylcholine chloride, and 1×10−6; 2×10−6; 5×10−6; 10×10−6 M of α-chaconine, respectively), which are measured during experiments. To verify the model and compare it with the experimental response a potentiometric biosensor based on immobilized butyrylcholine chloride was used. Selection of direct and inverse rate constants of enzymatic reactions was carried out in such a way that the result of numerical modeling corresponded as much as possible to the experimental response of the studied biosensor. A comparative analysis of the experimental and simulated responses of the biosensor for the determination of αchaconine was established. It was found that the absolute error does not exceed 0.045 units. As a result of computer simullation, it was concluded that the developed kinetic model of the potentiometric biosensor makes it possible to identify all the main components that were measured this study.

The full paper can be found here: https://ceur-ws.org/Vol-3468/paper1.pdf

The audio version is available:

The paper focuses on the challenges associated with deploying deep neural networks (DNNs) for the recognition of traffic objects using the camera of Android smartphones. The main objective of this research is to achieve resource-awareness, enabling efficient utilization of computational resources while maintaining high recognition accuracy. To achieve this, a methodology is proposed that leverages the Edge-to-Fog paradigm to distribute the inference workload across multiple tiers of the distributed system architecture. The evaluation was conducted using a dataset comprising real-world traffic scenarios and diverse traffic objects. The main findings of this research highlight the feasibility of deploying DNNs for traffic object recognition on resource-constrained Android smartphones. The proposed Edge-to-Fog methodology demonstrated improvements in terms of both recognition accuracy and resource utilization, and viability of both edge-only and edge-fog based approaches. Moreover, the experimental results showcased the adaptability of the system to dynamic traffic scenarios, thus ensuring real-time recognition performance even in challenging environments.

The link to conference can be found here: https://2023.euro-par.org/

The audio version is available: