Module 1: Foundations of AI: AI Technology Stack: Platforms, Applications and Model Types

Question 1

Define an AI Platform, describe what it can do and provide some examples.

Answer 1

Definition:
Software used to develop, test, deploy and refresh AI applications.

Capabilities:
- Centralize data analysis
- Streamline development and production workflows
- Facilitate collaboration
- Automate systems-development tasks
- Monitor models and systems in production

Examples:
- Google Cloud Platform
- Microsoft Azure
- Amazon Web Services

Question 2

Define an AI Application and provide some examples.

Answer 2

Defintion:
How an AI system is used.

Examples:
- Autonomous vehicles
- Chat bots
- E-commerce
- Education
- Facial recognition
- Finance
- Health care
- Human resources
- Marketing
- Navigation
- Robotics
- Social media

Question 3

What do Linear and Statistical Models do? What is an advantage of these models? Provide an example.

Answer 3

They model the relationship between 2 variables.

Advantage: They are not black box models and are therefore more explainable.

Example:
Linear regression model used to show how sales of a product are related to pricing based on historical data.

Question 4

What do Decision Trees Models do? What are their advantages and disadvantages?

Answer 4

Predict an outcome based on a flowchart of questions and answers.

Advantage: They are not a black box and are therefore more explainable.

Disadvantages:
- Changing even a little bit of the training data can have a significant impact on the algorithm.
- They are subject to security attacks and hacks.

Question 5

What are the disadvantages of Machine Learning Models? Provide an example.

Answer 5

Their black box capabilities make transparency and explainability more difficult.

Example: Neural Networks

Question 6

What is a Neural Network? What is a use case?

Answer 6

A Machine Learning Model that contains nodes in a layered structure and continuously improves the ability to find the right answer. They do not need to be trained to make complex non-linear inferences in unstructured data.

Use case: Facial recognition

Question 7

What are the different types of neural networks?

Answer 7

• Computer vision models (used to recognize images and videos)
• Speech recognition models (e.g. Alexa, transcription software)
• Language models (e.g. customer service chatbots)
• Reinforcement learning models