Advanced

Pattern Recognition Guide

This is your reference card. Every problem in this course maps to one of 8 core patterns. Learn to identify the pattern in the first 2 minutes of reading a problem, and you will solve it faster than someone who has memorized the solution but does not understand the pattern.

The Pattern Decision Tree

When you read a new problem, ask these questions in order:

Complete Problem-to-Pattern Map

Here is every problem in this course mapped to its primary pattern, with the key insight that unlocks the solution.

Pattern 1: Hash Map / Set

#	Problem	Key Insight
1	Two Sum	Store complement in map; check before inserting
10	Contains Duplicate	Set membership check is O(1)
15	Group Anagrams	Sorted string or char count tuple as hash key
18	Subarray Sum Equals K	Prefix sum + hash map of frequencies

Pattern 2: Two Pointers

#	Problem	Key Insight
3	Merge Two Sorted Lists	Compare heads, advance the smaller one
5	Valid Palindrome	Compare from both ends, skip non-alphanumeric
6	Linked List Cycle	Fast/slow pointers meet if cycle exists
11	3Sum	Fix one element, two pointers on sorted remainder
12	Container With Most Water	Start widest, shrink from shorter side
43	Trapping Rain Water	Track left_max and right_max from both ends

Pattern 3: Sliding Window

#	Problem	Key Insight
13	Longest Substring Without Repeating	Expand right, shrink left on duplicate
47	Sliding Window Maximum	Monotonic deque maintains decreasing order
50	Minimum Window Substring	Expand to satisfy, shrink to minimize

Pattern 4: Stack

#	Problem	Key Insight
2	Valid Parentheses	Push openers, pop and match closers
37	Min Stack	Auxiliary stack tracks min at each level
48	Largest Rectangle in Histogram	Monotonic increasing stack; pop on shorter bar

Pattern 5: Tree / Graph Traversal (BFS & DFS)

#	Problem	Key Insight
8	Binary Tree Inorder	Left-Root-Right with stack or recursion
21	Level Order Traversal	BFS with queue, process one level at a time
22	Validate BST	DFS with valid range (min, max) passed down
23	Number of Islands	DFS flood fill on each unvisited land cell
24	Course Schedule	Topological sort detects cycles in DAG
25	Word Search	Backtracking DFS with visited marking
26	Clone Graph	DFS/BFS + hash map from original to clone
27	Binary Tree Paths	DFS with path string building
28	Kth Smallest in BST	Inorder traversal gives sorted order; stop at k
29	Serialize/Deserialize Tree	Preorder with null markers for reconstruction
30	Lowest Common Ancestor	Recursive DFS; node is LCA if both subtrees return non-null
46	Word Ladder	BFS on implicit graph of one-letter changes

Pattern 6: Dynamic Programming

#	Problem	Key Insight
7	Climbing Stairs	dp[i] = dp[i-1] + dp[i-2] (Fibonacci variant)
9	Maximum Subarray	Kadane's: current_sum = max(num, current_sum + num)
31	Coin Change	dp[amount] = min coins; try each coin denomination
32	Word Break	dp[i] = can s[0:i] be segmented; check all splits
33	House Robber	dp[i] = max(skip, rob + dp[i-2])
34	Unique Paths	dp[i][j] = dp[i-1][j] + dp[i][j-1]
35	Decode Ways	Like climbing stairs but with digit constraints
44	Longest Palindromic Substring	Expand around each center (not classic DP table)
45	Edit Distance	2D DP: insert, delete, replace transitions
49	Regular Expression Matching	2D DP: handle '.' and '*' transitions carefully

Pattern 7: Binary Search

#	Problem	Key Insight
41	Median of Two Sorted Arrays	Binary search on partition of shorter array

Pattern 8: Heap / Greedy

#	Problem	Key Insight
38	Top K Frequent Elements	Bucket sort by frequency or min heap of size k
39	Meeting Rooms II	Min heap of end times; reuse room if available
40	Task Scheduler	Greedy: most frequent task determines structure
42	Merge K Sorted Lists	Min heap of k list heads; always pop smallest

Other Patterns

#	Problem	Pattern	Key Insight
4	Best Time to Buy/Sell Stock	Track Min	Track minimum price seen; max profit = price - min
14	Product of Array Except Self	Prefix/Suffix	Left pass for prefixes, right pass for suffixes
16	Merge Intervals	Sort + Scan	Sort by start; merge if overlap with previous
17	Sort Colors	Three-way Partition	Dutch National Flag with three pointers
19	Rotate Array	Reverse	Reverse all, reverse first k, reverse rest
20	Spiral Matrix	Simulation	Shrink boundaries after each pass
36	LRU Cache	Design	Hash map + doubly linked list for O(1) operations

How to Identify Patterns Quickly

Here are the trigger words and phrases that point to each pattern:

Pattern	Trigger Words in Problem Statement
Hash Map	"find pair", "two numbers that", "group", "frequency", "count occurrences"
Two Pointers	"sorted array", "pair/triplet", "palindrome", "in-place", "cycle"
Sliding Window	"contiguous subarray", "substring", "window of size k", "longest/shortest"
Stack	"matching brackets", "nested", "next greater", "monotonic", "undo"
BFS/DFS	"shortest path", "connected components", "levels", "tree", "graph", "grid"
DP	"minimum/maximum", "number of ways", "can you reach", "subsequence", "partition"
Binary Search	"sorted", "find position", "minimize maximum", "O(log n)"
Heap	"top k", "k-th largest", "merge k", "schedule", "priority"

Pattern Frequency at AI/ML Companies

Based on interview data, here is how frequently each pattern appears:

Pattern	Frequency	Must-Know Problems
Hash Map / Set	Very High (25%)	Two Sum, Group Anagrams, Subarray Sum K
Tree / Graph	High (20%)	Number of Islands, Course Schedule, LCA
Dynamic Programming	High (18%)	Coin Change, Word Break, Edit Distance
Two Pointers	Medium (12%)	3Sum, Trapping Rain Water
Sliding Window	Medium (8%)	Longest Substring, Minimum Window
Design	Medium (7%)	LRU Cache, Min Stack
Heap	Medium (6%)	Merge K Lists, Top K Frequent
Stack	Low (4%)	Valid Parentheses, Largest Rectangle

Key Takeaways

• There are only 8 core patterns. Every problem is a variation of one (or occasionally two) patterns.
• Spend 2 minutes reading the problem and identifying the pattern before writing any code. This saves time in the long run.
• Hash maps and tree/graph traversal account for nearly half of all interview questions at AI/ML companies.
• When stuck, try the trigger word analysis: look for keywords in the problem statement that map to patterns.
• Print this page and use it as a reference card during practice sessions.