Note: Intro -- Preface -- Organization -- Contents -- Evaluating the Robustness of Question-Answering Models to Paraphrased Questions -- 1 Introduction -- 2 Methodology -- 2.1 BERT, GPT-2 and XLNet -- 2.2 Embeddings -- 2.3 SQuAD -- 2.4 Para-SQuAD -- 2.5 Para-SQuAD Sentence Embeddings -- 3 Results -- 3.1 Experiments on QA Models -- 3.2 Dev-Para Performance -- 3.3 Visualization of Para-SQuAD -- 4 Related Work -- 5 Conclusion -- 5.1 Discussion -- 5.2 Limitations -- References -- FlipOut: Uncovering Redundant Weights via Sign Flipping -- 1 Introduction -- 2 Method -- 2.1 Motivation -- 2.2 FlipOut: Applying the Aim Test for Pruning -- 3 Related Work -- 3.1 Deep-R -- 3.2 Magnitude and Uncertainty Pruning -- 4 Experiments -- 4.1 General Setup -- 4.2 Choosing the Hyperparameters for FlipOut -- 4.3 Comparison to Baselines -- 4.4 Is It Just the Noise? -- 5 Discussion -- References -- Evolving Virtual Embodied Agents Using External Artifact Evaluations -- 1 Introduction -- 1.1 Computational Aesthetic Evaluation -- 1.2 Embodiment -- 2 Related Work -- 3 Implementation -- 3.1 Agent Morphology -- 3.2 Agent Control Policy -- 4 Experiment -- 4.1 Setup -- 4.2 Measuring Aesthetic Quality -- 4.3 Training Procedure -- 5 Results -- 6 Discussion -- 6.1 Future Work -- 7 Conclusion -- References -- Continuous Surrogate-Based Optimization Algorithms Are Well-Suited for Expensive Discrete Problems -- 1 Introduction -- 2 Problem Description -- 3 Related Work -- 4 Benchmark Problems -- 5 Experiments -- 5.1 Results -- 6 Conclusion and Future Work -- References -- Comparing Correction Methods to Reduce Misclassification Bias -- 1 Introduction -- 2 Methods -- 2.1 Baseline Estimator - Random Sample -- 2.2 Classify and Count -- 2.3 Subtracting Estimated Bias -- 2.4 Misclassification Probabilities -- 2.5 Calibration Probabilities -- 3 Results. , 3.1 Sampling Distributions of the Estimators -- 3.2 Finding the Optimal Estimator -- 4 Conclusion and Discussion -- A Appendix -- References -- A Spiking Neuron Implementation of Genetic Algorithms for Optimization -- 1 Introduction -- 2 High-Level Architecture -- 3 Neural Ensembles for Genetic Algorithms -- 3.1 Sequential Design -- 3.2 Parallel Design -- 4 Analysis -- 5 Discussion -- References -- Solving Hofstadter's Analogies Using Structural Information Theory -- 1 Introduction -- 1.1 Hofstadter's Analogies -- 1.2 Structural Information Theory -- 2 Analogy Solving -- 2.1 Structural Compression -- 2.2 Generating Symbols from Symbols -- 2.3 Symbol Substitution -- 2.4 Structure in Parameters -- 2.5 Inversion Trick -- 3 Python Implementation of PISA -- 3.1 Configurable Complexity Metric -- 3.2 Other Differences with PISA -- 4 Results -- 4.1 Murena's Dataset -- 4.2 Our Dataset -- 4.3 Complexity Values -- 5 Discussion -- 6 Future Developments -- References -- A Semantic Tableau Method for Argument Construction -- 1 Introduction -- 2 Preliminaries -- 3 Basic Argumentation Tableau -- 4 Defeasible Rules -- 4.1 Predicate Logic -- 4.2 Correctness and Completeness -- 5 Reasoning by Cases -- 5.1 Cases in an Argumentation Tableau -- 5.2 How to Reason by Cases with Defeasible Information -- 5.3 Local Tableau Closures -- 5.4 Mutually Exclusive Cases -- 6 Conclusion -- References -- `Thy Algorithm Shalt Not Bear False Witness': An Evaluation of Multiclass Debiasing Methods on Word Embeddings -- 1 Introduction -- 2 Background -- 3 Methodology -- 3.1 Terminology -- 3.2 Bias Measurements Techniques -- 3.3 Debiasing Techniques -- 4 Analysis of Religious Bias in Word Embeddings -- 4.1 Data -- 4.2 Analysis -- 5 Experiments and Results -- 5.1 Experimental Setup -- 5.2 RNSB Metric on Word Embeddings -- 5.3 WEAT and MAC on Word Embeddings. , 5.4 SoftWEAT Hyperparameter Experimentation -- 6 Conclusion -- References -- An Intelligent Tree Planning Approach Using Location-Based Social Networks Data -- 1 Introduction -- 2 Related Work -- 3 Problem Definition -- 4 Methods -- 4.1 Policy 1 - Degree -- 4.2 Policy 2 - Tree Density -- 4.3 Policy 3 - Community Tree Density -- 4.4 Policy 4 - Pareto Ranking -- 5 Experimental Set-Up -- 5.1 Data Sources -- 5.2 The Number of Iterations of the Community Detection Algorithm -- 5.3 Simulation -- 5.4 Evaluation -- 6 Results -- 6.1 Initial Situation -- 6.2 Ranking the Venues -- 6.3 Planting the Trees and Analyzing the Result -- 7 Conclusion -- References -- Gaining Insight into Determinants of Physical Activity Using Bayesian Network Learning-9pt -- 1 Introduction -- 2 Preliminaries -- 2.1 Bayesian Network Model -- 2.2 Learning Bayesian Networks -- 2.3 Handling Missing Data -- 3 Description of the Data and Methodology -- 3.1 Data Acquisition and Description -- 3.2 Data Preprocessing and Concept Design -- 3.3 Missing Data Analysis -- 3.4 Approach -- 4 Results -- 4.1 Comparison Bayesian Network Missing Data Strategy -- 4.2 Comparison of Bayesian Network Model to Previous Analyses -- 5 Discussion and Conclusions -- 6 Future Work -- References -- Swarm Construction Coordinated Through the Building Material -- 1 Introduction -- 2 Construction Algorithms -- 2.1 Standard Algorithms -- 2.2 Block Algorithms -- 3 Experiments -- 3.1 Dynamic Construction Paths -- 3.2 Guided Construction -- 3.3 Flexible Construction -- 4 Discussion -- 4.1 Fault Tolerance -- 4.2 Trade-Offs -- 5 Conclusion -- References -- Correction to: Gaining Insight into Determinants of Physical Activity Using Bayesian Network Learning. , Correction to: Chapter "Gaining Insight into Determinants of Physical Activity Using Bayesian Network Learning" in: M. Baratchi et al. (Eds.): Artificial Intelligence and Machine Learning, CCIS 1398, https://doi.org/10.1007/978-3-030-76640-5_11 -- Author Index.