Lecture Notes: AI Bot and AI Lab

Introduction

• Speaker: Dan, curriculum writer at code.org
• Topic: How AI Bot helps with data-driven decisions using AI Lab

AI and Machine Learning Overview

• Typical AI/ML examples seen at decision-making stage
• Recent lessons: Starting the AI/ML process
  • Investigating data to find patterns
  • Using patterns to make decisions
• Key terms:
  • Features: Inputs for decisions
  • Model: Created system for making decisions
  • Label: Output from decisions

Human vs. Machine in Data Analysis

• Humans can analyze small data and find patterns
• Difficulties arise with large data amounts
• AI Bot:
  • Analyzes large datasets
  • Identifies patterns humans might miss
• Machine learning aids by analyzing and finding trends quickly

Human Role in AI/ML Process

• Humans decide:
  • Problems for AI Bot to solve
  • Data to use
  • Important features to focus on
• Human decisions impact AI predictions
• Post-predictions: Humans decide on actions based on AI output

Introduction to AI Lab

• Tool for working with AI Bot
• Capabilities:
  • Preparing data
  • Training machine learning models
  • Evaluating model performance
• Interface walkthrough:
  • Central screen: Data set view
  • Example: Shoe Survey Data

Building a Model in AI Lab

• Sentence starter for model predictions
• Selection of columns as features:
  • Detailed column information in the right panel
  • Crosstab tables for feature usefulness
  • Example analysis: Bed preference vs. Sandals
• Bar charts for value frequency in data
• Adding features to the model
  • Button to add features
  • Sentence updates with selected features
  • Option to remove features

Training the Model

• Press the train button to create and evaluate the model based on chosen features