Feature engineering is the process of selecting, manipulating, and transforming raw data into features that can be used in supervised learning. In order to make machine learning work well on new tasks, it might be necessary to design and train better features.
This paper examines the use of Large Language Models (LLMs) for the intralingual translation of documents from standard German to German Easy Language (Leichte Sprache). We use open-weight models, from the Llama 3 family, with less than ten billion parameters.… Read More »LLMs for Easy Language Translation: A Case Study on German Public Authorities Web Pages
Ziel dieser Arbeit ist die Entwicklung und Validierung eines automatisierten Prognosemodells für Gepäckmengen am Hamburger Flughafen unter Verwendung der Low-Code AutoML-Bibliothek PyCaret. Durch die Automatisierung signifikanter Phasen des Machine-Learning-Lebenszyklus konnten präzise Vorhersagen für Gepäckstücke pro Flug innerhalb und außerhalb der… Read More »Automatisierte prädiktive Analytik in der Gepäckabfertigung
This paper presents a reinforcement learning approach for foreign exchange trading. Inspired by technical analysis methods, this approach makes use of technical indicators by encoding them into Gramian Angular Fields and searches for patterns that indicate price movements using convolutional… Read More »Scalp the Foreign Exchange Market with Deep Reinforcement Learning
This paper implements a ROS package for segmenting Point Clouds using Gradual Point Cloud Processing (GPCP). The implementation is integrated with RTAB-Map to produce a segmented point cloud of the environment. The RGB-D SLAM Dataset was used to validate the… Read More »Point Cloud Segmentation using Projection: A ROS Implementation
Detecting erroneous or fraudulent business transactions and corre-sponding journal entries imposes a significant challenge for auditors during annualaudits. One possible solution to cope with these problems is the use of machinelearning methods, such as an autoencoder, to identify unusual journal… Read More »XAI in the Audit Domain – Explaining an Autoencoder Model for Anomaly Detection
Deep learning methods have gained attention in predictive process mining tasks. The most popular architectures include Recurrent Neural Networks and Convolutional Neural Networks. Inspired by the fact that Graph Neural Networks can operate directly on graph data, this work proposes… Read More »Predictive Business Process Monitoring using Graph Neural Networks
Mobile robot path planning is one of the most essential problems in robotics. This work proposes a novel PSO-based method for mobile robot local path planning in dynamic unknown environments. Using this method, the robot is able to accelerate or… Read More »Mobile Robot Path Planning in Dynamic Unknown Environments using Particle Swarm Optimization
This work introduces Multi-Fusion Network for human-object interaction detection with multiple cameras. We present a concept and implementation of the architecture for a beverage refrigerator with multiple cameras as proof-of-concept. We also introduce an effective approach for minimizing the required… Read More »Recognizing Human-Object Interaction in Multi-Camera Environments
This paper introduces a concept for a self-supervised approach to apply a pre-trained 2D model for processing 3D point clouds. This approach helps to eliminate the need for 3D labeled data, which is extremely labor-intensive to create. The approach is… Read More »Self-supervised gradual 3D point clouds processing for robots