Key points from AI Presentation

Ai Presentation Key Points

Artificial Intelligence -> Machine Learning -> Deep Learning

Machine Learning

• Supervised
• Un-Supervised
• Reinforcement

Supervised - Label Training

• Spam Filter
• Image Classification 
• Predicting Prices

Un-Supervised

• Customer Segmentation in marketing
• Credit Card Fraud
• Language Processing

Reinforcement

• Autonomous Cars
• Automated Warehouses

Deep Learning

• Convolutional Neural Networks - Image Based
• Recurrent Neural Networks - Language & Speech
• Generative Adversarial Network - Generative Images/Art

Data

Is raw facts & Statistics, is required for Ai. Crucial to have good reliable sources for the Ai to be effective

Data Forms

Structured: Numbers, Dates, Categories

Un-structured: Text, Images, Audio

The Ai systems should have a split of 80/20 of learning/testing, the ai should be capable of learning by itself so it is important not to over teach it.

Quality of Data

Data must be:

• Accurate
• Complete
• Relevant
• Structured
• Error Free
• Unbiased

Ai - Related to Project

- Relies on data

Input - What is needed & where did it come from?

Output - What is is & how might it look/sound?

Research Method 

What data & where does it come from?

What does it need to show?

What do we want to understand & know?

Understand what you don't know

- Offer a solution

- Ask questions

- Respond to feedback

Propose Solution & offer explanation

Problem - Data Input

Solution - Data Visualisation