Overview | Agenda | Register

Course Outline:

Day 1: 8:30-4:30pm


8:30 Welcome and Introductions

8:45 Module 1: Foundations of AI

  • The Promise of AI
  • Why does AI matter
  • AI Reality vs. AI Myth
  • A Brief History of AI
  • Why is AI a “thing” now?
  • Definitions of AI - Terminology and Definitions
  • Narrow vs. Strong AI
  • The Core Aspects of Intelligence
  • Machine Learning and Cognitive Technology
  • Supervised, Unsupervised, and Reinforcement Learning
  • Deep Learning: A Revolution in Neural Networks
  • Augmented Intelligence
  • Our AI-Enabled Vision of the Future
  • The DIKUW Pyramid (AI is Based on Data)
  • What is Data Science and why is it relevant to AI
  • Realizing the Promise of AI: The AI-Enabled Vision of the Future
  • Data-centric methodologies for AI
  • Leveraging CRISP-DM Methodology
  • Agile Methodology & CPMAI : Dual Iteration Loops
  • Introducing CPMAI

10:45 Module 2: Patterns of AI (Part 1)

  • Do We Even Need AI?
  • Where AI is Best Suited
  • The 7 Patterns of AI
  • Pattern: Conversations & Content
    • Conversational Use Cases
    • Conversational Industry Examples
    • Conversational Case Study
  • Pattern: Recognition
    • Recognition Use Cases
    • Recognition Industry Examples
    • Recognition Case Study
  • Pattern: Autonomous Systems
    • Autonomous Use Cases
    • Autonomous Industry Examples
    • Autonomous Case Study

12:00 LUNCH

13:00 Module 2: Patterns of AI (Part 2)

  • Pattern: Hypersonalization
    • Hyperpersonalization Use Cases
    • Hyperpersonalization Industry Examples
    • Hyperpersonalization Case Study
  • Pattern: Pattern & Anomaly Detection
    • Pattern & Anomaly Detection Use Cases
    • Pattern & Anomaly Detection Industry Examples
    • Pattern & Anomaly Detection Case Study
  • Pattern: Predictive Analytics & Decision Support
    • Predictive Analytics & Decision Support Use Cases
    • Predictive Analytics & Decision Support Industry Examples
    • Predictive Analytics & Decision Support Case Study
  • Pattern: Goal-Driven Systems
    • Goal-Driven Systems Use Cases
    • Goal-Driven Systems Industry Examples
    • Goal-Driven Systems Case Study
  • Tackling AI Projects as Combinations of Patterns
    • Examples with multiple patterns in a single project

14:45 Module 3: CPMAI Phase I: Business Understanding

  • ML Anti-Patterns
  • Case Studies in AI
  • Real-World Examples and Use Cases
  • Identifying Value Propositions for AI & Cognitive Tech
  • Developing Business-Centric ROI
  • Autonomous vs. Augmented / Assisted Intelligence Approaches
  • Probabilistic vs. Deterministic Systems
  • The AI Go/No-Go Decision
  • Scoping AI Projects
  • CPMAI Phase I Deliverables

16:30 Conclusion of Day 1

Day 2: 8:30-4:30pm


08:30 Module 4: CPMAI Phase II: Data Understanding

  • AI Projects Need a Foundation in Big Data
  • Moving our way up the DIKUW Pyramid
  • Structured, Unstructured, Semi-Structured data
  • What is Data Science?
  • What is Data Engineering?
  • Data Scientists vs. Data Engineers
  • Relationship between ML & Data Science
  • 80% of AI Projects are Data Engineering
  • What Did We Learn from Big Data?
  • Big Data Infrastructure
  • Big Data Methodologies
  • Applying what we learned from Big Data Projects
  • Do you even need Big Data for AI?
  • Taking a Data-Centric Mentality on AI
  • CPMAI Phase II Deliverables

10:45 Module 5: CPMAI Phase III: Data Preparation

  • Data engineering and preparation

12:00 LUNCH

13:00 Module 6: CPMAI Phase IV: Model Development (Part 1)

  • Intelligence requires learning
  • Machine Learning: A Technology Definition
  • How do we encode experience?
  • Machine Learning Terminology
  • ML Concept: Dimensions
  • The Curse of Dimensionality: Why Machine Learning is Hard
  • ML Concept: Classification
  • ML Concept: Regression
  • Diving Deeper into the Three Forms of Machine Learning
  • Supervised Learning
  • Unsupervised Learning
  • Reinforcement Learning
  • Machine Learning Algorithms & Models
  • Supervised Learning & ML Concepts Visually Explained
  • Finding the Line: Cost Functions, Gradient Descent and More
  • Neural Networks: Mimicking the Brain
  • Deep Learning: Making Neural Networks Work
  • ML Model Training: Tuning the Knobs
  • Flavors of Deep Learning Neural Networks
  • What exactly is (in) a ML model?
  • Transfer Learning: Speeding Things Up
  • The Pre-Trained Model
  • Using Pre-Trained Models (Networks)

14:45 Module 6: CPMAI Phase IV: Model Development (Part 2)

  • Other approaches to Supervised Learning
  • Support Vector Machines (SVM)
  • Decision Trees
  • Random Forests
  • Bayesian Classifiers
  • K-Nearest Neighbors
  • Machine Learning Challenge: The Bias-Variance Tradeoff
  • Resolving the Bias-Variance Tradeoff: Cross-Validation
  • Improving Performance with Ensemble Methods
  • Unsupervised Learning & ML Concepts Visually Explained
  • Neural Network approaches to Unsupervised Learning
  • K-Means Clustering
  • Gaussian Mixture Models
  • t-SNE Example
  • Autoencoders and GANs
  • Reinforcement Learning & ML Concepts Visually Explained
  • Q-Learning
  • DeepMind Examples
  • The Limits of Machine Learning
  • Machine Reasoning & Common Sense
  • Knowledge Graphs
  • AI is still evolving

16:30 Conclusion of Day 2

Day 3: 8:30-4:30pm


08:30 Module 7: CPMAI Phase V: Model Evaluation

  • Model Evaluation & Testing
  • The Confusion Matrix
  • Business / KPI evaluation
  • Model iteration

10:45 Module 8: CPMAI Phase VI: Model Operationalization (Pt. 1)

  • AI Vendor Classification
  • Big Data infrastructure and AI
  • The four “platforms” of AI
  • ML as a Service and Cloud ML
  • AI operationalization
  • The Data Science Environment
  • The Data Science Notebook
  • Data Science Notebook Choices: Jupytr, Colab, and More
  • Setting up a Good Data Notebook Environment
  • Data Visualization and AI
  • Python, R, MATLAB, Mathematica, and More
  • ML Frameworks / Toolkits Compared: Keras, TensorFlow, Caffe(2), Pytorch, MXNet, and more
  • Model Development: Python Scikit-Learn
  • Model Training
  • Accelerating Model Training: GPUs, TPUs, and More
  • AutoML

12:00 LUNCH

13:00 Module 8: CPMAI Phase VI: Model Operationalization (Pt. 2)

  • Vendors in the Data Science Environment
  • Pre-Trained Models and Networks
  • The Big Data & Data Engineering Environment
  • Hadoop & Spark
  • Data Preparation Tools
  • Data Labeling Solutions
  • Vendors in the Data Engineering Environment
  • Model “Scaffolding” Environment
  • Machine Learning as a Service (MLaaS)
  • Cloud ML Services
  • Cloud Training vs. Cloud Inference
  • Point-Solution AI Providers: Recognition
  • Point-Solution AI Providers: Predictive Analytics & Decision Support
  • Point-Solution AI Providers: Conversation & Content Generation
  • Point-Solution AI Providers: Industry-Specific
  • Point-Solution AI Providers: Patterns & Anomalies
  • Vendors in the “Middle” AI Environment
  • The Operational Environment
  • Operationalizing ML at the “Edge”
  • Operationalizing ML in the Server
  • Operationalizing ML in the Cloud
  • The Problem with Pseudo AI

14:45 Module 9: Ensuring Responsible AI

  • Keeping the human in the loop
  • Addressing fears and concerns about AI
  • Transparent & explainable AI (XAI)
  • Emerging laws and regulations
  • Addressing issues of bias and human induced error
  • Tackling Fears of AI
  • Tackling Real Concerns about AI
  • AI is not a Job Killer, but it is a Job Category Killer
  • Automation & The Amazon Paradox
  • Making AI Transparent & Explainable
  • Detecting Informational Bias in Datasets
  • AI Introduces new Threat Vectors: Malicious AI
  • AI Can Make the Fake Seem Real… and the reverse
  • The “Uncanny Valley”
  • Privacy in an Era of AI-Enhanced Big Data
  • Pervasive surveillance
  • The rise of Adversarial attacks on Computer Vision systems
  • Concerns over Artificial General Intelligence (AGI)
  • The “Singularity”
  • Resolving these Issues: Keep Humans in the Loop
  • China: The Future or Dystopia?
  • AI Laws & Regulations
  • Organizational AI Ethics & Governance
  • Resolving these Issues: Provide Transparency
  • GDPR and AI Transparency
  • Resolving these Issues: Prioritize Fairness and Dignity
#AIWORLD