CSC345 Content

White Board
What is AI? (R&N Ch.1)
- Historical answers
- Moving Answer
- Your answers?
- R&N answers: Traditional; Agentic;
- My answers: Philosophical; Computational (example); Structural.
Sensors
- Knowledge sources: Human (we think we know it all); Environment (temperature, sights, sounds, animals, ocean, etc.); Machine (data loggers, IoT, machine learning, etc.)
- Source structures: Raw data (sampling, data logs, etc.); Formal representations (math, engineering, science, etc.); Informal representations (natural language, images, sounds, etc.)
- Data Collecting (Data Science)
- Natural Language Processing (R&N Ch.23-24)
  - Overview (View from IBM)
  - Symbolic NLP
    Symbolic NLP in Python (explained, with part of the book, the API documentation)
  - NLP for LLMs (during LLMs)
  - Language to Logic (after Knowledge Representation)
Knowledge Representation
- The Status of Knowledge
- Vector Data
- Knowledge graphs (R&N Ch.10)
  - Views from IBM; neo4j; Wikipedia
  - Knowledge graphs using NLP
    Knowledge graphs in Python (explained).
  - Knowledge graphs vs. RDF triples vs. Property graphs: Views from neo4j.
  - Property graphs in Python (explained).
- Logic (R&N Ch.8-9)
  - The TPTP World
  - Classical logic: Propositional; First-order; Typed First-order
  - Non-classical logic
- Taxonomies and Ontologies
  - Taxonomy views from Wikipedia; YAGO
  - Ontology views from Adam Pease (SEoP); Adam Pease (talk); SUMO (Adam's ontology)
Reasoning with Knowledge
- K-nearest neighbours
  KNN classification in Python, KNN regression in Python
- Search (R&N Ch.3-4)
- Search in application (R&N Ch.5-6)
  - Constraints
  - Game search (slides)
- Reasoning in logic (R&N Ch.7-9)
  - ATP; Logical consequence; LC in PRP by truth tables;
  - CNF; Resolution; ANL loop;
  - Examples in the TPTP World: PRP; FOL; TFF
- Uncertainty and Imprecision (R&N Ch.12-13)
  - Uncertainty (slides)
    - Bayes' rule (more slides; even more slides)
  - Imprecision;
- Genetic algorithms (R&N Ch.4.1.4)
Learning from Knowledge, and Reasoning Again
- Types of learning (slides)
- Evaluation of supervised learning
- K-Means (video)
- Decision trees (slides)
- Linear regression
  Linear regression in Python
- Neural networks: notes, slides; Back propagation: notes, slides
- Large language models
  - LLM Prompt engineering
  - Taxonomies in LLMs
Actuators
- Robotics
- Natural language generation
  - Views from IBM
Ethics
- For AI
- In Academia
Conclusion