Intro
Preface
Organization
Contents
AI System Engineering: Math, Modelling and Software
Unboundedness of Linear Regions of Deep ReLU Neural Networks
1 Introduction
2 Preliminaries
3 Representation and Analysis via Code Space
4 Experiments
5 Outlook and Future Work
References
Applying Time-Inhomogeneous Markov Chains to Math Performance Rating
1 Introduction
1.1 The Story Behind
Origins of the Use Case
1.2 An Adaptive Mobile App for Math Training
2 Ranking Based on the Elo System

2.1 Automated Estimation of Skill and Difficulty as a Problem-Solving Match Between Learner and Exercise
2.2 Time-inhomogeneous Markov Chains
3 Data and Experiment
4 Analysis and Results
5 Conclusions
Practical Use of the Ranking System
References
A Comparative Analysis of Anomaly Detection Methods for Predictive Maintenance in SME
1 Introduction
1.1 Anomaly Detection Techniques:
2 Literature Review
3 Methodology
3.1 Data
3.2 Model Description
3.3 Test Description
4 Results
4.1 Performance Comparison
5 Conclusion
References

A Comparative Study Between Rule-Based and Transformer-Based Election Prediction Approaches: 2020 US Presidential Election as a Use Case
1 Introduction
2 Related Work
3 Methodology
3.1 Data Collection
3.2 Pre-processing and Filtering
3.3 Sentiment Analysis
3.4 Vote-Share
4 Results and Discussion
4.1 Sentiment Analysis
4.2 Vote-Share
5 Conclusion
References
Detection of the 3D Ground Plane from 2D Images for Distance Measurement to the Ground
1 Introduction
2 Related Work
2.1 Depth Estimation
2.2 Plane Detection

3 Extraction of 3D Ground Plane Equation from RGB Images
3.1 Depth Estimation
3.2 3D Ground Plane Detection
4 Experiments and Results
4.1 3D Ground Plane Detection from RGB Images (Entire Workflow)
4.2 3D Ground Plane Detection from Groundtruth Depth Maps (Intermediate Workflow)
4.3 Influence of Bounding Point Selection on Ground Plane Detection
5 Conclusion and Outlook
References
Towards Practical Secure Privacy-Preserving Machine (Deep) Learning with Distributed Data
1 Introduction
1.1 The State-of-Art

1.2 Requirements for a Practical Secure Privacy-Preserving Machine Learning
2 Proposed Methodology
3 A Biomedical Application Example
4 Concluding Remarks
References
Applied Research, Technology Transfer and Knowledge Exchange in Software and Data Science
Collaborative Aspects of Solving Rail-Track Multi-sensor Data Fusion
1 Introduction
2 Problem Statement
3 Collaborative Problem Solving Approach
3.1 Roles and Competences
3.2 Collaborative Development of a Calibration Object
3.3 Feedback-driven, Iterative Data Recording
3.4 Project Results