Description

Goals
What might be the goals of the intelligent system?

○ Assist students in learning how to do math problems
■ Using this assistance to then walkthrough where exactly they went wrong within the problem
■ Also will allow the student tho then practice the types of problems that they get wrong
○ Provide adaptive feedback
■ Adapt to the student’s style of learning and level of knowledge
■ Award the student when they are correct
■ Explain and walkthrough with the student when they are wrong
■ Track student progress, provide problems, and assessments
○ Increase student engagement and motivation
■ Ensure the student is comfortable with the section of math they are working on
■ With some students that struggle with mathematics, it can involve them feeling very stressed. We want to make sure the student feels as if they are in a stress-free environment so we strive for the student to not feel stressed.
■ Increased accessibility for students to help them in developing their math skills.
Environment
What is the environment that the system will be adapting within? What are the important characteristics of that environment?
● Learning/Classroom Environment
○ Characteristics to consider: system should mirror aspects of physical classroom(i.e. Lesson structure, topic division/categories).
Kids have short attention spans so needs to draw their focus
● Mathematical problem sets
○ Characteristics to consider: large number of students with different levels of understanding and struggle with different tasks
■ Possible solution could be a standard intro test to gauge where the students’ math skills are at.
● Multiple choice questions and fill in the blanks
○ Characteristics to consider: wide variety of answer possibilities with fill in the blank questions
Adaptation
How might the system adapt to the environment given the goals? How might its inner environment change?

Teachers use a variety of tools to teach students and better explain concepts, why shouldn’t this proposed system? A future version of this system might be able to provide explanations of the steps for solving that specific problem as well as a more general ChatGPT explanation for the subject or specific type of problem
● Example:
○ Our system output:
■ 2*2=4 -> “add the first number to itself the second number times” ○ ChatGPT:
■ “Multiplication is a mathematical operation that involves combining two or more quantities to produce a third quantity that represents the total value of the combined quantities…”
● Decision Tree Model
○ Breaking down given problems into subproblems
● Neural Network
○ Understand the student’s inputs and provide personalized feedback
○ Using a dataset of math problems could train the model
● Reinforced Learning (RL)
○ Allow the model to adapt and improve its response using feedback