Explore the intersection of artificial intelligence and symbolic reasoning through this comprehensive 12-hour tutorial series covering fundamental concepts in neuro-symbolic AI. Master both generative and discriminative AI approaches while diving deep into System 1 and System 2 thinking paradigms. Learn logic programming, open world reasoning, and propositional logic alongside modern language models, prompt engineering, and encoder-decoder architectures. Understand probabilistic graphical models, Signal Temporal Logic (STL), and gradient descent optimization techniques. Examine the technical shortcomings of deep learning through detailed analysis and discover how logical inference and first-order logic can address these limitations. Study Markov Random Fields, Markov Chains, Markov Logic Networks, and Bayesian reasoning principles. Gain practical experience with backpropagation, transformer architectures, activation functions, convolution operations, RNNs, embeddings, and deep reinforcement learning. Explore machine learning operations (MLOps), distance functions, and Probabilistic Soft Logic (PSL) applications. Delve into unsupervised learning techniques including k-means clustering, hierarchical clustering, and cluster evaluation methods. Master rule learning, itemset mining, association rule mining, and the A Priori algorithm. Understand semi-supervised and self-supervised learning approaches while learning experimental design principles for AI and ML research. Get hands-on experience with PyReason through practical tutorials covering pet store examples, binary classifiers, and time series reasoning integration. Conclude with guidance on generating innovative AI research ideas and conducting meaningful experiments in the field.

Syllabus

Generative vs. Discriminative AI
Thinking Fast and Thinking Slow: System 1 and System 2
Introduction to Logic Programming and Open World Reasoning
Introduction to Propositional Logic
Introduction to Language Models (LLM's, Prompt Engineering, Encoder/Decoder and more)
Probabilistic Graphical Models
STL: Signal Temporal Logic
Introduction to Gradient Descent
Technical Shortcoming of Deep Learning, pt. 1 of 2
Technical Shortcoming of Deep Learning, pt. 2 of 2
Introduction to Logic (for AI)
Inference in propositional logic
Tutorial: First Order Logic (FOL, Predicate Calculus)
Markov Random Fields, Markov Chains, Markov Logic Networks, and more
Introduction to Bayesian Reasoning
Training Challenges in Deep Learning
Backpropagation: Training Deep Neural Networks
The difference between AI and machine learning
Understanding the factors changing the distribution of data
Machine Learning and Discrimination
Basic Concepts in Bias and Fairness in Machine Learning
Discussion on select advances in neuro symbolic reasoning and learning
Regularization in Machine Learning
Transformer Architecture
Activation Functions used in Deep Neural Networks
Kernels and the Convolution Operation
Understanding RNNs
Embeddings
Deep Reinforcement Learning
Introduction to ML Ops
Distance Functions
Introduction to Probabilistic Soft Logic (PSL)
Applications of Reinforcement Learning
Parametric vs Non Parametric Approaches
k Nearest Neighbor
Unsupervised Learning and Clustering
k-Means Clustering
Hierarchical Clustering
Evaluating and Selecting Unsupervised Learning Methods
Introduction to Rule Learning and Itemset Mining
Itemset Mining and the A Priori Algorithm
Association Rule Mining
Semi-Supervised and Self-Supervised Learning
Ideas on Conducting Experiments in AI and ML
Quick and Dirty Intro to Neurosymbolic AI
Intro PyReason Tutorial: Pet Store Example
Advanced PyReason Tutorial
Pt I: PyReason - ML integration tutorial (binary classifier)
Pt II: PyReason - ML integration tutorial (time series reasoning)
Creating AI Ideas for Research