UMass Lowell Computer Science Analysis of Algorithms Prof. Karen Daniels Fall, 2004 Final Review

Review of Key Course Material

What’s It All About? ä Algorithm: ä steps for the computer to follow to solve a problem ä Problem Solving Goals: ä recognize structure of some common problems ä understand important characteristics of algorithms to solve common problems ä select appropriate algorithm & data structures to solve a problem ä tailor existing algorithms ä create new algorithms

Some Algorithm Application Areas Computer Graphics Geographic Information Systems Robotics Bioinformatics Astrophysics Medical Imaging Telecommunications Design Apply Analyze

Tools of the Trade ä Algorithm Design Patterns such as: ä binary search ä divide-and-conquer ä randomized ä Data Structures such as: ä trees, linked lists, stacks, queues, hash tables, graphs, heaps, arrays Growth of Functions Summations Recurrences Sets Probability MATH Proofs

Discrete Math Review Growth of Functions, Summations, Recurrences, Sets, Counting, Probability

Topics ä Discrete Math Review : ä Sets, Basic Tree & Graph concepts ä Counting: Permutations/Combinations ä Probability: Basics, including Expectation of a Random Variable ä Proof Techniques: Induction ä Basic Algorithm Analysis Techniques: ä Asymptotic Growth of Functions ä Types of Input: Best/Average/Worst ä Bounds on Algorithm vs. Bounds on Problem ä Algorithmic Paradigms/Design Patterns: ä Divide-and-Conquer, Randomized ä Analyze pseudocode running time to form summations &/or recurrences

What are we measuring? ä Some Analysis Criteria: ä Scope ä The problem itself? ä A particular algorithm that solves the problem? ä “Dimension” ä Time Complexity? Space Complexity? ä Type of Bound ä Upper? Lower? Both? ä Type of Input ä Best-Case? Average-Case? Worst-Case? ä Type of Implementation ä Choice of Data Structure

Function Order of Growth O( ) upper bound  ( ) lower bound  ( ) upper & lower bound n 1 n lg(n) n lg 2 (n) 2n2n2n2n n5n5n5n5 lg(n) lg(n)lglg(n) n2n2n2n2 know how to use asymptotic complexity notation to describe time or space complexity know how to order functions asymptotically (behavior as n becomes large) shorthand for inequalities

Types of Algorithmic Input Best-Case Input: of all possible algorithm inputs of size n, it generates the “best” result for Time Complexity: “best” is smallest running time for Time Complexity: “best” is smallest running time Best-Case Input Produces Best-Case Running Time Best-Case Input Produces Best-Case Running Time provides a lower bound on the algorithm’s asymptotic running time provides a lower bound on the algorithm’s asymptotic running time (subject to any implementation assumptions) (subject to any implementation assumptions) for Space Complexity: “best” is smallest storage for Space Complexity: “best” is smallest storage Average-Case Input Worst-Case Input these are defined similarly Best-Case Time <= Average-Case Time <= Worst-Case Time

Bounding Algorithmic Time (using cases) n 1 n lg(n) n lg 2 (n) 2n2n2n2n n5n5n5n5 lg(n) lg(n)lglg(n) n2n2n2n2 T(n) =  (1) T(n) =  (2 n ) very loose bounds are not very useful! Worst-Case time of T(n) =  (2 n ) tells us that worst-case inputs cause the algorithm to take at most exponential time (i.e. exponential time is sufficient). But, can the algorithm every really take exponential time? (i.e. is exponential time necessary?) If, for arbitrary n, we find a worst-case input that forces the algorithm to use exponential time, then this tightens the lower bound on the worst-case running time. If we can force the lower and upper bounds on the worst-case time to match, then we can say that, for the worst-case running time, T(n) =  (2 n ) (i.e. we’ve found the minimum upper bound, so the bound is tight.) Using “case” we can discuss lower and/or upper bounds on: best-case running time or average-case running time or worst-case running time

Bounding Algorithmic Time (tightening bounds) n 1 n lg(n) n lg 2 (n) 2n2n2n2n n5n5n5n5 lg(n) lg(n)lglg(n) n2n2n2n2 T W (n) =  (2 n ) for example... Here we denote best-case time by T B (n); worst-case time by T W (n) T B (n) =  (1) 1st attempt T B (n) =  (n) 1st attempt 2nd attempt T W (n) =  (n 2 ) T B (n) =  (n) 2nd attempt 1st attempt T W (n) =  (n 2 ) Algorithm Bounds

ä Explore the problem to gain intuition: ä Describe it: What are the assumptions? (model of computation, etc...) ä Has it already been solved? ä Have similar problems been solved? (more on this later) ä What does best-case input look like? ä What does worst-case input look like? ä Establish worst-case upper bound on the problem using an algorithm ä Design a (simple) algorithm and find an upper bound on its worst-case asymptotic running time; this tells us problem can be solved in a certain amount of time. Algorithms taking more than this amount of time may exist, but won’t help us. ä Establish worst-case lower bound on the problem ä Tighten each bound to form a worst-case “sandwich” Approach increasing worst-case asymptotic running time as a function of n n 1 2n2n2n2n n2n2n2n2 n3n3n3n3 n4n4n4n4 n5n5n5n5

Know the Difference! n 1 2n2n2n2n n5n5n5n5 worst-case bounds on problem on problem An inefficient algorithm for the problem might exist that takes this much time, but would not help us. No algorithm for the problem exists that can solve it for worst-case inputs in less than linear time. Strong Bound: This worst-case lower bound on the problem holds for every algorithm that solves the problem and abides by our problem’s assumptions. Weak Bound: This worst-case upper bound on the problem comes from just considering one algorithm. Other, less efficient algorithms that solve this problem might exist, but we don’t care about them! Both the upper and lower bounds are probably loose (i.e. probably can be tightened later on).

Master Theorem MMaster Theorem : LLet with a > 1 and b > 1. Tthen : CCase 1: If f(n) = O ( n (log b a) -  ) for some  > o Tthen T ( n ) =  ( n log b a ) CCase 2: If f (n) =  (n log b a ) Tthen T ( n ) =  (n log b a * log n ) CCase 3: If f ( n ) =  (n ( log b a) +  ) for some  > o and if a f( n/b) N 0 Tthen T ( n ) =  ( f ( n ) ) Use ratio test to distinguish between cases: f(n)/ f(n)/ n log b a Look for “polynomially larger” dominance.

Master Theorem

CS Theory Math Review Sheet The Most Relevant Parts... ä p. 1  O, ,  definitions ä Series ä Combinations ä p. 2 Recurrences & Master Method ä p. 3 ä Probability ä Factorial ä Logs ä Stirling’s approx ä p. 4 Matrices ä p. 5 Graph Theory ä p. 6 Calculus ä Product, Quotient rules ä Integration, Differentiation ä Logs ä p. 8 Finite Calculus ä p. 9 Series Math fact sheet (courtesy of Prof. Costello) is on our web site.

Sorting Chapters 6-9 Heapsort, Quicksort, LinearTime-Sorting

Topics ä Sorting: Chapters 6-8 ä Sorting Algorithms: ä [Insertion & MergeSort)], Heapsort, Quicksort, LinearTime-Sorting ä Comparison-Based Sorting and its lower bound ä Breaking the lower bound using special assumptions ä Tradeoffs: Selecting an appropriate sort for a given situation ä Time vs. Space Requirements ä Comparison-Based vs. Non-Comparison-Based

Heaps & HeapSort ä Structure: ä Nearly complete binary tree ä Convenient array representation ä HEAP Property: (for MAX HEAP) ä Parent’s label not less than that of each child ä Operations: strategy worst-case run-time ä HEAPIFY: swap downO(h) [h= ht] ä INSERT: swap upO(h) ä EXTRACT-MAX: swap, HEAPIFY O(h) ä MAX: view rootO(1) ä BUILD-HEAP: HEAPIFY O(n)  HEAP-SORT: BUILD-HEAP, HEAPIFY  (nlgn)

QuickSort ä Divide-and-Conquer Strategy ä Divide: Partition array ä Conquer: Sort recursively ä Combine: No work needed ä Asymptotic Running Time:  Worst-Case:  (n 2 ) (partitions of size 1, n-1)  Best-Case:  (nlgn) (balanced partitions of size n/2)  Average-Case:  (nlgn) (balanced partitions of size n/2) ä Randomized PARTITION ä selects partition element randomly ä imposes uniform distribution Does most of the work on the way down (unlike MergeSort, which does most of work on the way back up (in Merge) PARTITION Recursively sort right partition right partition left partition Recursively sort left partition

Comparison-Based Sorting In algebraic decision tree model, comparison-based sorting of n items requires  (n lg n) worst-case time. HeapSort To break the lower bound and obtain linear time, forego direct value comparisons and/or make stronger assumptions about input. InsertionSort MergeSort QuickSort  (n)  (n 2 ) BestCaseAverageCaseWorstCase Time: Algorithm:  (n lg n)   (n lg n)  (n lg n)*  (n lg n)  (n lg n)  (n lg n)  (n 2 ) (*when all elements are distinct)

Data Structures Chapters Stacks, Queues, LinkedLists, Trees, HashTables, Binary Search Trees, Balanced Trees

Topics ä Data Structures: Chapters ä Abstract Data Types: their properties/invariants ä Stacks, Queues, LinkedLists, (Heaps from Chapter 6), Trees, HashTables, Binary Search Trees, Balanced (Red/Black) Trees ä Implementation/Representation choices -> data structure ä Dynamic Set Operations: ä Query [does not change the data structure] ä Search, Minimum, Maximum, Predecessor, Successor ä Manipulate: [can change data structure] ä Insert, Delete ä Running Time & Space Requirements for Dynamic Set Operations for each Data Structure ä Tradeoffs: Selecting an appropriate data structure for a situation ä Time vs. Space Requirements ä Representation choices ä Which operations are crucial?

Hash Table ä Structure: ä n << N (number of keys in table much smaller than size of key universe) ä Table with m elements ä m typically prime ä Hash Function: ä Not necessarily a 1-1 mapping ä Uses mod m to keep index in table ä Collision Resolution: ä Chaining: linked list for each table entry ä Open addressing: all elements in table ä Linear Probing: ä Quadratic Probing: Load Factor: Example:

Linked Lists ä Types ä Singly vs. Doubly linked ä Pointer to Head and/or Tail ä NonCircular vs. Circular ä Type influences running time of operations / head9 4 3 / head tail9 4 3 head

Binary Tree Traversal ä “Visit” each node once  Running time in  (n) for an n-node binary tree ä Preorder: ABDCEF ä Visit node ä Visit left subtree ä Visit right subtree ä Inorder: DBAEFC ä Visit left subtree ä Visit node ä Visit right subtree ä Postorder: DBFECA ä Visit left subtree ä Visit right subtree ä Visit node B E C F D A

Binary Search Tree B D F E A C ä Structure: ä Binary tree ä BINARY SEARCH TREE Property: ä For each pair of nodes u, v: ä If u is in left subtree of v, then key[u] <= key[v] ä If u is in right subtree of v, then key[u] >= key[v] ä Operations: strategy worst-case run-time ä TRAVERSAL: INORDER, PREORDER, POSTORDER O(h) [h= ht] ä SEARCH: traverse 1 branch using BST property O(h) ä INSERT: search O(h) ä DELETE: splice out (cases depend on # children) O(h) ä MIN: go left O(h) ä MAX: go right O(h) ä SUCCESSOR: MIN if rt subtree; else go up O(h) ä PREDECESSOR: analogous to SUCCESSOR O(h) ä Navigation Rules ä Left/Right Rotations that preserve BST property

Red-Black Tree Properties ä Every node in a red-black tree is either black or red ä Every null leaf is black ä No path from a leaf to a root can have two consecutive red nodes -- i.e. the children of a red node must be black ä Every path from a node, x, to a descendant leaf contains the same number of black nodes -- the “black height” of node x. newly inserted node

Graph Algorithms Chapter 22 DFS/BFS Traversals, Topological Sort

Topics ä Graph Algorithms: Chapter 22 ä Undirected, Directed Graphs ä Connected Components of an Undirected Graph ä Representations: Adjacency Matrix, Adjacency List ä Traversals: DFS and BFS ä Differences in approach: DFS: LIFO/stack vs. BFS:FIFO/queue ä Forest of spanning trees ä Vertex coloring, Edge classification: tree, back, forward, cross ä Shortest paths (BFS) ä Topological Sort ä Tradeoffs: ä Representation Choice: Adjacency Matrix vs. Adjacency List ä Traversal Choice: DFS or BFS

Introductory Graph Concepts: Representations B E C F D A B E C F D A ä Undirected Graph ä Directed Graph (digraph) A B C D E F ABCDEF ABCDEF A BC B ACEF C AB D E E BDF F BE A BC B CEF C D D E BD F E Adjacency Matrix Adjacency List Adjacency Matrix Adjacency List

SEARCHING Elementary Graph Algorithms: SEARCHING: DFS, BFS ä Breadth-First-Search (BFS): ä BFS  vertices close to v are visited before those further away  FIFO structure  queue data structure ä Shortest Path Distance ä From source to each reachable vertex ä Record during traversal ä Foundation of many “shortest path” algorithms See DFS, BFS Handout for PseudoCode ä Depth-First-Search (DFS): ä DFS backtracks  visit most recently discovered vertex  LIFO structure  stack data structure ä Encountering, finishing times : “well- formed” nested (( )( ) ) structure ä DFS of undirected graph produces only back edges or tree edges ä Directed graph is acyclic if and only if DFS yields no back edges for unweighted directed or undirected graph G=(V,E) Time: O(|V| + |E|) adj listO(|V| 2 ) adj matrix predecessor subgraph = forest of spanning trees Vertex color shows status: not yet encountered encountered, but not yet finished finished

Elementary Graph Algorithms: DFS, BFS ä Review problem: TRUE or FALSE? ä The tree shown below on the right can be a DFS tree for some adjacency list representation of the graph shown below on the left. B E C F D A A C B E D F Tree Edge Cross Edge Back Edge

Elementary Graph Algorithms: Topological Sort source: textbook Cormen et al. TOPOLOGICAL-SORT(G) 1 DFS(G) computes “finishing times” for each vertex 2 as each vertex is finished, insert it onto front of list 3 return list for Directed, Acyclic Graph (DAG) G=(V,E) Produces linear ordering of vertices. For edge (u,v), u is ordered before v. See also DFS/BFS slide show