Expert systems compared with machine learning

Name:

Date:

Class:

Quick Questions

Shade in the A or B circle for each question.

1. Who writes the rules in an expert system?

A Experts and people put them in

B The system writes them by itself

2. Which type of AI adapts its own processes and data?

A Expert system

B Machine learning

3. How does the knowledge base of an expert system get updated?

A People add or change the facts

B The expert system rewrites it automatically

4. Which statement is correct?

A Both expert systems and machine learning are types of AI

B Only machine learning counts as AI

5. A program follows IF-THEN rules written by doctors to suggest a diagnosis. What is it?

A Machine learning

B An expert system

Match the Pairs

Draw a line connecting each item on the left to its match on the right.

Left column		Right column
1IF symptom = fever AND rash THEN measles		AWrong because both are types of AI. They differ in how they work: experts' rules versus adapting by itself.
2accuracy_new > accuracy_old: keep change		Bknowledge_base.add(fact_from_expert)
3An expert system's knowledge base only changes when people update it		Cdata.add(new_example) # done by the program
4Machine learning adapts its own data by itself as it gets more examples		DA rule written by a person (an expert) and stored in an expert system's rule base
5Expert system: the program rewrites its own rules as it learns		EMachine learning testing a change to its own process and keeping it if results improve
6Machine learning is AI but an expert system is not AI		FWrong because an expert system does not adapt itself. Its rules and facts are put in by experts. Adapting by itself is machine learning.

Definitions

Expert system vs machine learning

Fill in the Blanks

Complete each line of code by filling in the missing parts.

rules ← getRulesFrom(_______) result ← infer(facts, rules)

model._____(newData) model.save()

knowledgeBase.______(newFact) knowledgeBase.save()

IF type = "expert" __ type = "ML" THEN isAI ← TRUE ENDIF

IF adaptsOwnData = ____ THEN type ← "machine learning" ENDIF

Spot the Error

Shade in the letter of the line that contains an error.

Question 1

A expert_system:

B rules <- set by experts

C facts <- set by experts

D is_AI <- FALSE

Question 2

A ML: needs experts to adapt

B ML: adapts own data

C ES: uses rules from experts

Question 3

A IF system uses expert rules

B THEN type <- expert system

C ELSE IF system adapts itself

D THEN type <- expert system

Question 4

A INPUT new_fact FROM expert

B ES adds new_fact by itself

C OUTPUT updated expert system

Question 5

A ML program gets new data

B Expert rewrites its process

C ML result improves

Answer Key

Expert systems compared with machine learning

Section A: Quick Questions

A — Experts and people put them in

B — Machine learning

A — People add or change the facts

A — Both expert systems and machine learning are types of AI

B — An expert system

Section B: Match the Pairs

IF symptom = fever AND rash THEN measles → A rule written by a person (an expert) and stored in an expert system's rule base

accuracy_new > accuracy_old: keep change → Machine learning testing a change to its own process and keeping it if results improve

An expert system's knowledge base only changes when people update it → knowledge_base.add(fact_from_expert)

Machine learning adapts its own data by itself as it gets more examples → data.add(new_example) # done by the program

Expert system: the program rewrites its own rules as it learns → Wrong because an expert system does not adapt itself. Its rules and facts are put in by experts. Adapting by itself is machine learning.

Machine learning is AI but an expert system is not AI → Wrong because both are types of AI. They differ in how they work: experts' rules versus adapting by itself.

Section C: Definitions

Expert system vs machine learning — An expert system reasons using rules and facts put in by experts. Machine learning adapts its own processes and data by itself. Both are types of AI.

Section D: Fill in the Blanks

rules ← getRulesFrom(experts)
result ← infer(facts, rules)

model.adapt(newData)
model.save()

knowledgeBase.update(newFact)
knowledgeBase.save()

IF type = "expert" OR type = "ML" THEN
  isAI ← TRUE
ENDIF

IF adaptsOwnData = TRUE THEN
  type ← "machine learning"
ENDIF

Section E: Spot the Error

D — An expert system is a type of AI, so is_AI should be TRUE.

A — Machine learning adapts its own processes by itself. It does not need experts to change it. Line 1 should say that ML adapts its own processes.

D — A system that adapts itself is machine learning (ML), not an expert system. The line should be THEN type <- ML.

B — An expert system does not update itself. People add new facts to the knowledge base. The line should add the expert's fact to the knowledge base.

B — Machine learning changes its own process by itself. An expert does not have to rewrite it.