AI 2023: Advances in Artificial Intelligence : 36th Australasian Joint Conference on Artificial Intelligence, AI 2023, Brisbane, QLD, Australia, November 28–December 1, 2023, Proceedings, Part II / edited by Tongliang Liu, Geoff Webb, Lin Yue, Dadong Wang.

Format:

Book

Contributor:

Liu, Tongliang, editor.

Series:

Lecture Notes in Artificial Intelligence, 2945-9141 ; 14472

Language:

English

Subjects (All):

Artificial intelligence.

Computer networks.

Data mining.

Application software.

Computer vision.

Artificial Intelligence.

Computer Communication Networks.

Data Mining and Knowledge Discovery.

Computer and Information Systems Applications.

Computer Vision.

Local Subjects:

Artificial Intelligence.

Computer Communication Networks.

Data Mining and Knowledge Discovery.

Computer and Information Systems Applications.

Computer Vision.

Physical Description:

1 online resource (509 pages)

Edition:

1st ed. 2024.

Place of Publication:

Singapore : Springer Nature Singapore : Imprint: Springer, 2024.

Summary:

This two-volume set LNAI 14471-14472 constitutes the refereed proceedings of the 36th Australasian Joint Conference on Artificial Intelligence, AI 2023, held in Brisbane, QLD, Australia during November 28 – December 1, 2023. The 23 full papers presented together with 59 short papers were carefully reviewed and selected from 213 submissions. They are organized in the following topics: computer vision; deep learning; machine learning and data mining; optimization; medical AI; knowledge representation and NLP; explainable AI; reinforcement learning; and genetic algorithm.

Contents:

Intro

Preface

Organization

Contents - Part II

Contents - Part I

Knowledge Representation and NLP

Collaborative Qualitative Environment Mapping

1 Introduction

2 Qualitative Spatio-Temporal Reasoning

3 LH Interval Calculus

4 Collaborative Qualitative Environmental Mapping

5 Experiments

6 Related Work

7 Conclusion and Future Work

References

Towards Learning Action Models from Narrative Text Through Extraction and Ordering of Structured Events

2 Related Work

3 Structured Event Extraction

4 Event Ordering

5 Narrative Chain Extraction

6 Challenges for NLP Research

7 Challenges for Model Acquisition

8 Conclusions

The Difficulty of Novelty Detection and Adaptation in Physical Environments

2 Background and Related Work

2.1 Novelty Research

2.2 Difficulty Prediction

2.3 Learning Algorithms

2.4 Qualitative Spatial Relations (QSRs)

2.5 Experimental Domain

3 Novelty Difficulty Formulation

3.1 Dimensions of Novelty

3.2 Observational State

3.3 Action State

4 Discussion and Conclusion

Lateral AI: Simulating Diversity in Virtual Communities

1.1 Large Language Models

1.2 Prompt Engineering

2 Lateral AI

2.1 Lateral AI Design

2.2 Comparison with Other Models

2.3 Key Features of Lateral AI

3 Lateral AI Demonstrations

3.1 Arnold Schwarzenegger AI Persona's Advice on Vitality

3.2 Creating Unconventional Thinkers

3.3 A Moral Dilemma

3.4 Seeking Recommendations from a Board of Experts

3.5 Pushing AI To Predict Beyond Its Factual Knowledge

4 Conclusion

Reports, Observations, and Belief Change

2 Preliminaries

2.1 Motivating Example

2.2 Belief Revision

2.3 Trust

3 Revision by Reports.

3.1 Basic Definitions

3.2 Report Revision Operators

3.3 Honesty Sets

3.4 Representation Result

4 Observations

4.1 Conflict

4.2 Revision by Observations

4.3 Basic Properties

5 Discussion

5.1 Related Work

5.2 Future Work

5.3 Conclusion

A Prompting Framework to Enhance Language Model Output

2 Prompting Techniques

3 Research Methods

3.1 Framework Formulation

4 Results

4.1 Experiments

4.2 Intrinsic Evaluation Results

4.3 Extrinsic Evaluation Results

4.4 Constraints

5 Conclusions

Epistemic Reasoning in Computational Machine Ethics

2 Background

3 Ethical Principle Function

3.1 Goodness-Based Principle

3.2 Less Harm Principle

3.3 Deontological Principle

4 Aggregation Strategies

4.1 Maximum Average

4.2 Maximin Strategy

4.3 Coefficient of Optimism

4.4 Regret Minimisation

4.5 Illustration

5 Results and Discussions

5.1 Milnor's Axioms

5.2 Axiom Satisfaction

6 Conclusion

Using Social Sensing to Validate Flood Risk Modelling in England

2 Methodology

2.1 Data Collection

2.2 Twitter Data Pre-processing

2.3 Flood Map Development

3 Results

4 Discussion

5 Conclusion

Symbolic Data Analysis to Improve Completeness of Model Combination Methods

2.1 The Symbolic Data Analysis Paradigm

2.2 Consensus Models

3 Build a Decision Tree from a Symbolic Data Table

3.1 Build a Decision Tree from a Synthetic Dataset

3.2 Build a Decision Tree from Symbolic Distributional Data

4 Evaluation and Results

4.1 Datasets

4.2 Results and Discussion

CySpider: A Neural Semantic Parsing Corpus with Baseline Models for Property Graphs

1 Introduction.

2 Related Work

3 Notation and Task Formulation

4 SQL2Cypher: From SQL Queries to Cypher Queries

5 Text-to-Cypher Neural Models

5.1 Pipeline

5.2 End-to-End Training

5.3 Evaluation Metric

6 Experiment Results

6.1 Dataset Statistics

6.2 Models Evaluation Result

6.3 Error Analysis

S5TR: Simple Single Stage Sequencer for Scene Text Recognition

3 Proposed Method

4 Experiments

4.1 Datasets and Evaluation Metrics

4.2 Implementation Details

4.3 Comparisons with the State-of-the-arts

4.4 Qualitative Results

Explainable AI

Coping with Data Distribution Shifts: XAI-Based Adaptive Learning with SHAP Clustering for Energy Consumption Prediction

2.1 Energy Consumption Prediction

2.2 XAI-Based Model Improvement

2.3 SHapley Additive ExPlanations (SHAP)

3 SHAP Clustering-Based Adaptive Learning (SCAL)

3.1 Building Block 1: SHAP Clustering in Explanation Space

3.2 Building Block 2: Extraction of SHAP Clustering Characteristics

3.3 Building Block 3: Adaptive Model Refinement Based on SHAP Clustering Characteristics

4 Experimental Setup and Data Set

5 Results

5.1 SCAL Performance

5.2 Cluster Analysis in Explanation Space

6 Transferability to Other Use Cases

6.1 Financial Distress Data Set (Classification Problem)

6.2 Power Data Set (Regression Problem)

Concept-Guided Interpretable Federated Learning

2.1 Interpretable Federated Learning

2.2 Concept-Related Interpretability

3 Problem Settings

3.1 Federated Learning

3.2 Concept Bottleneck Model

4 Proposed Method

4.1 Concept Bank

4.2 Linear Predictor.

4.3 Training Algorithm

5 Experiment

5.1 Datasets

5.2 Performance Analysis

5.3 Reasoning Process

6 Conclusion and Limitations

Systematic Analysis of the Impact of Label Noise Correction on ML Fairness

3 Methodology

6 Discussion

7 Conclusions

Part-Aware Prototype-Aligned Interpretable Image Classification with Basic Feature Domain

2 Method

2.1 The Overview of PaProtoPNet Architecture

2.2 Basic Feature Domain and Prototype Alignment

2.3 Feature Separation Module

2.4 Computing Scores for Classification

2.5 Overall Loss Function

3 Experiments

3.1 Performance Comparison

3.2 Model Analysis

3.3 Reasoning Process

4 Discussion and Future Work

Hybrid CNN-Interpreter: Interprete Local and Global Contexts for CNN-Based Models

2.1 Convolutional Neural Network Structures

2.2 Visual Interpretabilities of CNN-Based Models

3 Method

3.1 Stacking Forward Propagation

3.2 Linear Regression Module

3.3 Filter Importance Analysis Module

4 Experiment and Discussion

4.1 Local Interpretability for CNN-Based Models

4.2 Global Interpretability for CNN-Based Models

5 Conclusion and Future Work

Impact of Fidelity and Robustness of Machine Learning Explanations on User Trust

2.1 User Trust

2.2 Fidelity

2.3 Robustness

3 Hypotheses

4 Methodology

4.1 Fidelity-Based Scenario Study Design

4.2 Robustness-Based Scenario Study Design

4.3 Metrics

5.1 Dataset

5.2 Participants

5.3 Experimental Procedure

6 Results

6.1 Correlations Between User Trust and Fidelity

6.2 Correlations Between User Trust and Robustness

7 Discussion.

8 Conclusion and Future Work

Interpretable Drawing Psychoanalysis via House-Tree-Person Test

2 Related Works

2.1 Drawing Psychoanalysis

2.2 Class Activation Mapping

3.1 Quantization of the Size

3.2 Quantization of the Position

4.1 The HTP Dataset

4.2 Implementation Detials

4.3 Experiment Results

A Non-asymptotic Risk Bound for Model Selection in a High-Dimensional Mixture of Experts via Joint Rank and Variable Selection

1.1 Main Contributions

2 Collection of Polynomial SGaBloME Models

2.1 Variable Selection via Selecting Relevant Variables

2.2 Variable Selection via Rank Sparse Models

2.3 Collection of Polynomial SGaBloME Models

3 Main Theoretical Results

3.1 Boundedness Conditions on the Parameter Space

3.2 Loss Function

3.3 Penalized Maximum Likelihood Estimation (PMLE)

3.4 Oracle Inequality

4 Conclusion and Perspectives

Reinforcement Learning

Auction-Based Allocation of Location-Specific Tasks

2 Setup

3 Auction-Based Algorithms

3.1 Bidding Rule

3.2 BidSumPath Bidding Rule

3.3 BidSumTree Bidding Rule

4 Theoretical Analysis Under Task Capacities

5 Experimental Comparison of Algorithms

5.1 Experimental Setup and Design

5.2 Impact of Feasibility Constraints

5.3 Performance Against Optimal

6 Conclusions

Generalized Bargaining Protocols

2 Automated Negotiation

3 Proposed Framework

3.1 Evaluating Negotiation Protocols

3.2 Tentative Agreements Unique Offers (TAU)

4 Empirical Evaluation

SAGE: Generating Symbolic Goals for Myopic Models in Deep Reinforcement Learning

3 Related Work

4 SAGE.

4.1 Meta-controller.

Other Format:

Print version: Liu, Tongliang AI 2023: Advances in Artificial Intelligence

ISBN:

981-9983-91-6