Please enable JavaScript.
Coggle requires JavaScript to display documents.
AI Methods - Coggle Diagram
AI Methods
Machine Learning
What is the main difference between traditional programming and machine learning?
Traditional programming follows fixed rules; machine learning learns patterns from data.
Define "supervised training"
A learning method where the model is trained using labeled data.
How does the computer learn in "unsupervised training"?
The computer learns patterns or structures in data without labeled answers.
What is a 'cluster' in the context of unsupervised training?
A group of similar data points identified in unsupervised learning.
Explain what "reinforcement learning" involves.
Learning through trial and error by receiving rewards or penalties.
What does the text suggest is a unique feature of deep learning?
It uses layers of artificial neural networks to automatically extract features.
List two advantages of machine learning.
Can process large data quickly
Improves accuracy over time
What are some limitations of machine learning?
Requires lots of data.
Can be biased if data is biased.
Name two uses of machine learning.
Face recognition.
Spam detection.
How can machine learning help with weather forecasting?
It analyzes vast climate data to predict weather patterns more accurately.
Decision Tree
What is a 'decision tree'?
A decision tree is a way of laying out the different decisions. By following the branches of the tree, you can get to the right decision
Why are decision trees useful?
Because you can find the answer to your questions
Can decision trees be automated?
Yes, to do this you have to turn your decision tree into a program. Each branch of the tree represents an "if" "else" structure
Expert System
What are the components of an algorithm?
Inputs: data used by the algorithm
Processes: changes made to the data
Structures: such as loops and if structures
Outputs: the values produced by the algorithm
How does an algorithm help in making human-type decisions and judgments?
The algorithm will take all the factors that a human would consider. It will output a decision that matches what humans would design
What is an expert system and who does it represent?
An expert system is an algorithm that represents the knowledge of an expert
It represents:
Doctors
Judge
Engineer
What are some advantages and disadvantages of using an expert system?
Advantages:
You can use it when a person is not available
It will always make the same decision without bias or favour
Its rules are set out in program code so you can always check if they are fair
Disadvantages
It can be hard for an expert to put their skills into word
Not everything can be reduced to logical tests
Experts can be wrong, and the algorithm will be wrong too
Euristic