lab2_notes.txt

initial state:
(0, 0) = 0 && (m, n) = 0 => by convention
(0 0 0 0 .... 0
 0 0 0 0 .... 0
 .
 .
 0 0 0 0 .... 0)

transition:

(1 0 0 0 .... 0
 0 0 0 0 .... 0
 .
 .
 0 0 0 0 .... 0)

transition:

(1 1 0 0 .... 0
 0 0 0 0 .... 0
 .
 .
 0 0 0 0 .... 0)

final state:

(1 1 0 0 .... 0
 0 1 0 0 .... 0
 .
 .
 0 0 1 1 .... 1) - a matrix representing valid a path from (0,0) to (m,n)

(0, 0) = 1 && (m, n) = 1



MAZE GENERATION:
mxn matrix with possible values: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15
1 - North
2 - South
4 - East
8 - West

e.g.: 7 = 1 + 2 + 4 => North, South, East are free to explore
      9 = 1 + 8 => North, West are free to explore
=> this means it would be very easy to create a tree which sould explore all the possible routes

BACKTRACKING:
	- create the tree and explore all the possible routes in order (in depth)

BFS:
	- create the tree and explore all the possible routes in breadth-wise order

HILLCLIMBING:
	- h(n) = abs(Xfininsh - Xcurrent) + abs(Yfinish - Ycurrent)
	- h(n') = abs(Xfininsh - Xtarget) + abs(Yfinish - Ytarget)
 
SIMULATED ANNEALING:
	- start with a score of mxn (i.e. for n=5 and m=6, initial score=30)
	- loss applies if i am forced to make a "bad" move (all possible following states are "worse")
	- loss value for h(n') < h(n) is h(n)-h(n')
	- probability = e^(-loss/score)