Introduction to Artificial Intelligence and Machine Learning

Objectives

• Define artificial intelligence (AI)
• Describe the relationship between AI and data science
• Define machine learning (ML)
• Describe the relationship between ML, AI, and data science
• Describe different ML approaches
• Identify applications of ML

Emergence of AI

• Data Economy: Growth of data has led to AI development.
  • Since 2009, data volume increased by 44 times due to social websites.
  • Companies battle for data ownership.

Need for AI

• Big Data: AI helps manage and analyze massive volumes of data.
• Data Science: Provides insights by teaching machines to learn.

What is Artificial Intelligence?

• Definition: Intelligence displayed by machines simulating human and animal intelligence using logic.
• Functions: Sensing, reasoning, and acting.
• Applications:
  • Self-driving cars
  • Apple Siri
  • Google's AlphaGo
  • Amazon Echo
  • IBM Watson
• In Media: AI concepts widely featured in sci-fi movies.

AI in Commerce

• Recommendation Systems: Used by e-commerce to suggest products based on user behavior.

Relationship Between AI, ML, and Data Science

• AI: Mimics human intelligence.
• ML: Enables systems to learn and improve without explicit programming.
• Data Science: Encompasses data analytics, mining, ML, AI.
• Process Flow:
  1. Data Gathering and Transformation
  2. Predictions using ML techniques
  3. Insights from predictions
  4. Perform actions using AI

Machine Learning (ML)

• Relationship with AI: ML gives machines the ability to gain intelligence.
• Relationship with Data Science: ML uses data insights to form algorithms.

Features of Machine Learning

• Pattern Detection: Classifies data based on learned patterns using reinforcement learning.
• Algorithmic Learning: Learns from previous data for decision making.

Machine Learning Approaches

• Traditional vs. ML Programming:
  • Traditional: Hard-coded decision rules.
  • ML: Trains a model with data to derive algorithms.

Machine Learning Techniques

• Classification: Predicts discrete responses.
• Categorization: Organizing data into categories.
• Clustering: Grouping similar data objects.
• Trend Analysis: Projects event movements using time series data.
• Anomaly Detection: Identifies unusual data cases.
• Visualization: Presents data pictorially for decision making.

Real-Time Applications of Machine Learning

• Image Processing: Enhances images, e.g., Facebook tagging, OCR.
• Robotics: Humanoid emotion reading, industrial manufacturing.
• Data Mining: Fraud detection, market analysis.
• Video Games: Predicts outcomes based on data.
• Text Analysis: Spam filtering, sentiment analysis.
• Healthcare: Disease diagnosis, drug discovery, medical imaging.

Companies using ML in Healthcare

• Google DeepMind Health
• BioBeats
• Health Fidelity
• Ginger.io