D.6: AutoML Tool

• Short Description: The AutoML Tool is designed for domain users who are not experts in machine learning. It provides an easy-to-use, wizard-like interface for covering every step of data-driven machine learning. Users can upload data, create ad hoc queries, and perform preprocessing tasks such as normalization and standardization. It allows users to train and test machine learning algorithms, compare models using various metrics, and save the best-performing models. The goal is to make machine learning approachable and useful for operators and managers in the process industry, enabling them to use AI for process optimization without requiring in-depth technical knowledge of data science. The AutoML Tool is a framework that includes a graphical user interface for data preprocessing, model training, and testing. It supports data upload from databases or CSV files and uses customizable filters for data selection. Metrics such as F1 score, Area Under the Curve, and confusion matrices are used for model comparison. The system likely adheres to relevant data standards and uses common protocols and APIs for data handling and model deployment. AutoML Tool's optimization algorithms create a closed-loop system. This allows for continuously refining the model and control strategy based on real-time data feedback. The AutoML Tool supports relevant standards in ISO/IEC.

• Reference, URL: https://www.teknopar.com.tr/

• Applicable Business Category: Manufacturing

• Application in relevant Projects/Initiatives:Yers - Circular TwAIn

• Asset Type: Executable

• AI Breadth: ML

• Learning Ability: N.A.

• Related technologies: N.A

• Applicable Research Area: Collaborative AI

• Applicable Technical Category: Machine Learning

• License Information: N.A

• Related to circularity and sustainability: Yes

• Audience: Operators and Factory Managers

Go back to the AI portfolio of reference implementations