The increasing adoption of Artificial Intelligence (AI) in industrial environments has created a pressing need for efficient and effective methods to train, test, and compare AI models. Industrial applications, such as predictive maintenance, quality control, and process optimization, rely heavily on high-performing AI models to drive business decisions. However, developing and deploying these models can be a complex and time-consuming process, requiring significant expertise and computational resources. An application for training AI models and comparing their performance is essential to simplify this process. Such an application enables the rapid development, training, and evaluation of AI models on large datasets, enabling a more in-depth understanding and facilitating the identification of the best-performing model for a specific task. Moreover, a built-in opportunity to compare AI models allows for the evaluation of different algorithms, hyperparameters, and training data, providing valuable insights into model strengths and weaknesses, supporting effective human-AI cooperation. By streamlining AI model development and evaluation, industrial companies can reduce the time and cost associated with model deployment, improve model accuracy and reliability, and increase the overall efficiency of their AI-driven operations. This paper presents such a streamlined application, whereat the novelty is set on the operator friendly user-interface to enable the analysis, model training and evaluation for non-experts directly at the plant.
Luftensteiner et al. (Thu,) studied this question.