Scribd
Upload
69 views25 pages

Dynamic Programming (II) : Anson Ho

This document provides an overview of different types of dynamic programming problems including DAG DP, dimension reduction techniques, tree DP, and bitwise DP. It discusses key concepts in dynamic programming like states, transition formulas, base cases, and memorization. It provides examples of dynamic programming problems involving DAGs, trees, grids, bitmasks, assignments, matching, paths, and more. The document is intended to explain the different categories of dynamic programming problems and common approaches.

Uploaded by

sherlock holmes
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
69 views25 pages

Dynamic Programming (II) : Anson Ho

This document provides an overview of different types of dynamic programming problems including DAG DP, dimension reduction techniques, tree DP, and bitwise DP. It discusses key concepts in dynamic programming like states, transition formulas, base cases, and memorization. It provides examples of dynamic programming problems involving DAGs, trees, grids, bitmasks, assignments, matching, paths, and more. The document is intended to explain the different categories of dynamic programming problems and common approaches.

Uploaded by

sherlock holmes
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 25

Dynamic Programming (II)

Anson Ho
Contents
• DAG DP
• Dimension reduction
• Tree DP
• Bitwise DP
What is DP
• State(s)
• Transition formula
• Base case(s)

• Memorization

• Time complexity
• Space complexity
DAG
• Directed acyclic graph

• Visualization of transition
DAG
• Horner's method

• Fibonacci number

• Number of sequences of strictly


including rectangles with integral
dimensions
DAG
• Number of ways to pay $10

%!
• Calculate 𝑛𝐶𝑟 = '! %(' !

• Shortest path
Dimension reduction
• Rolling array

• Reduce dimension for space


• Not for time
Dimension reduction
• Number of paths in grids
– From top-left to bottom right

• 0: Only ↓ and → are allowed


• 1: Some blocked grids
• 2: 5000x5000 grid
• 2: 5000 blocked 1MB memory
• 3: output “median” path
Dimension reduction
• Avoid moving a large chunk of memory

• dp[i][j]
• = f(dp[i-1][j],dp[i-2][j+1])
• Only keep the last two dp[i]

• i à i mod 3
• i-1 à i-1 mod 3
• i-2 à i-2 mod 3
Take a break
Tree DP
• What is tree?

• A Tree is a DAG
• (A DAG is not necessarily a tree)

• Rooted or not?
Tree DP
• Rooted

• Height of each node

• Size of each subtree

• Minimax in a game tree


Tree DP
• Not rooted
– bidirectional edges

• Number of paths passing through


each node

• Sum of path weights


– path weight = sum of edge weights
– path weight = product of edge weights
Tree DP
• Number of “k-subtrees” in unrooted
tree

• “k-subtree” means a subgraph which


is a tree with k nodes
Tree DP
• Some of the nodes in a tree are
coloured

• For every node, find the number of


paths with coloured ends passing
through it
(Extra)
• Every node in a tree is coloured

• Queries with a specific node and a


specific colour
• Find the number of paths with ends
in that colour passing through
that node
Tree DP
• Given two rooted tree

• Find the minimum number of


additional nodes to make the two
trees “isomorphic”

• Assignment problem
Bitmask DP
• State
• dp[133][2] à dp[1011001012]

• Bit manipulation
– bitwise and (&)
– bitwise or (|)
– exclusive or (^)
Bitmask DP
• Assignment problem
• Matching

• Maximum matching
• Matching with minimum cost

• Polynomial time algorithm exists


Bitmask DP
• Hamiltonian path

• Graph traversal with visiting each


node exactly once
Bitmask DP
• N (<=15) light bulbs
• K (<=30) buttons
– each control a set of light bulbs

• Find the minimum number of


toggling to achieve a specific
configuration
Bitmask DP
• Number of ways to fill a NxM grid
with 1x2 (or 2x1) rectangle
– fully filled
– no overlapping

• N=2, M=2
• Ans = 2
Bitmask DP
• Define 2, 8, 9 as lucky digit
• Define numbers containing only
lucky digits as lucky number

• Find the least lucky multiple of


369
Bitmask DP
• 29889 = 369 x 81

• State 1: cur mod 369


• State 2: bitmask of used
• State 2: lucky digit
• DP from leftmost digit

• Ref: HackerEarth Lucky Digit


Thank you

You might also like