Lecture 19: Problem Solving Flashcards
Mental Operators
Applied to the initial state, get to a new state, apply more operators, and so on searching for the goal state
Well structured
stages are well defined (math)
○ You know what the initial state, goal state, and mental operators are
Ill Structured
stages are not clear (choosing a career)
Stages in Problems, Solving
○ Representing the problem
§ Very important, using graphs may make a problem easier or harder
○ Planning the problem
○ Executing the plan
○ Checking if it work
○ Retrying if needed
Buddhist Monk Example
- Monk travels up a mountain, a few days later travels back down, is there a spot on the mountain the monk will be at on the same spot as another monk at the same time?
○ Easier to solve with a visual representation; of course the monks will cross paths- Useful to retrieve a similar representation of the problem you are solving
Ray Problem
Figure out how to apply radiation to a tumor; the solution is to shoot weak waves in many directions that all converge
* Same thing works in a military problem, but it is hard to find the analogy between the two and solve one problem after the other
* It helps to give a hint to connect the problem; people can use relevant analogies but often fail to notice them
Top-Down Preconceptions
What you already know can sometimes hinder problem solving
Functional Fixedness
People assume the function of some object is fixed
* In the candle problem you have a tack, candle, and tack box, and most people don’t consider using the tack box because it is a box not a support * Same thing in the two-rope problem, people do not think about tying the pair of pliers to one of the ropes
Trains meeting problem
This bird flies back and forth between two trains driving towards each other. People well versed in calculus tried to solve using that but there is a much easier way to solve it
Luchin’s Water Jar Problem
People solve a task where you have to fill up bottles of water and get a specific quantity by filling it up and pouring out etc. and if you solve 5 or 6 problems in a row using the same sequence then people will stick with that strategy when it no longer applies, this is them being stuck in a mental set
Algorithms
The way you might program a computer; a sequence of steps that is guaranteed to produce a correct answer but requires a lot of way
* Unscrambling a word, guessing a password
Heuristics
Shortcuts that are not guaranteed but are fast and easy
Difference Reduction Heuristic
The idea of choosing the mental operator; if you have a lot you choose the one that reduces the difference the most between the goal and initial state
* Does not always work; sometimes we have to take a step back
Chess study
Expertise is pretty domain specific; being good at chess does not mean they are smart
* Experts are better at remembering chess boards that make sense, where novices are better at the opposite
Means-End Analysis Heuristic
Analyze the end and see what means may allow you to achieve that end
- Identify biggest difference between initial and goal state
- Do operator subgoaling; find the operator that will reduce the difference and satisfy the preconditions
- Flight, plan, airport, money, ticket, etc.