[ University Logo ] U N D E R G R A D U A T E C O U R S E · A R T I F I C I A L I N T E L L I G E N C E Introduction to Artificial Intelligence Concepts · Problem Solving · Search Techniques · Heuristics · Problem Reduction Faculty: [ Faculty Name ] Department: [ Department ] 10 - Hour Teaching Module · 40 Slides · For B.Tech / B.E. / BCA / MCA M O D U L E 1 · O R I E N T A T I O N Course Learning Outcomes Introduction to Artificial Intelligence 2 / 40 On completing this course, students will be able to: 1 Explain core AI concepts, characteristics and techniques 2 Analyse AI problem - solving approaches and problem spaces 3 Apply uninformed search — BFS and DFS — to state spaces 4 Use heuristic search: hill climbing and best - first search 5 Model and solve Constraint Satisfaction Problems (CSP) 6 Apply problem reduction and Means - Ends Analysis Course Roadmap 1 AI Concepts & Foundations 2 Problems & Problem Spaces 3 BFS & DFS Search 4 Heuristic Search 5 Reduction & CSP M O D U L E 1 · F O U N D A T I O N S What is Artificial Intelligence? Introduction to Artificial Intelligence 3 / 40 Definition Artificial Intelligence is the branch of computer science concerned with building systems that perform tasks normally requiring human intelligence — reasoning, learning, perception, language understanding and decision - making — by representing knowledge and searching for solutions. • Coined by John McCarthy at the 1956 Dartmouth Conference • Turing Test (1950) proposed a behavioural test for intelligence • Expert - system era (1980s) → statistical ML (2000s) → deep learning & generative AI (2010s – today) • Modern AI powers search, vision, speech, robotics and large language models 1950 Turing Test 1956 AI coined 1997 Deep Blue 2011 Watson / Siri 2022+ Generative AI M O D U L E 1 · F O U N D A T I O N S Goals and Objectives of AI Introduction to Artificial Intelligence 4 / 40 1 Human - like Intelligence Reproduce perception, language and reasoning so machines interact naturally with people. e.g. Chatbots, virtual assistants 2 Rational Agents Build agents that perceive the environment and act to maximise expected performance. e.g. Thermostats, game - playing AI 3 Problem Solving Represent problems as states and search for action sequences that reach a goal. e.g. Route planning, puzzles 4 Learning & Decision Making Improve from data and experience to make better, adaptive decisions over time. e.g. Fraud detection, recommendations M O D U L E 1 · F O U N D A T I O N S Characteristics of AI Introduction to Artificial Intelligence 5 / 40 1 Learning Acquiring knowledge and skills from data and experience. 2 Reasoning Drawing logical inferences to reach conclusions or decisions. 3 Knowledge Representation Encoding facts, rules and relations a machine can use. 4 Perception Interpreting sensory input — vision, speech, signals. 5 Adaptability Adjusting behaviour as the environment changes. 6 Autonomy Acting and deciding with little or no human intervention. M O D U L E 1 · F O U N D A T I O N S AI Techniques Introduction to Artificial Intelligence 6 / 40 • Knowledge Representation • Logic, semantic nets, frames, production rules • Search Techniques • Uninformed (BFS/DFS) and heuristic search • Machine Learning • Learning patterns from data • Expert Systems • Rule - based reasoning over a knowledge base • Reasoning Methods • Deduction, induction, abduction AI Techniques Knowledge Representation Search Techniques Machine Learning Expert Systems Reasoning Methods M O D U L E 1 · F O U N D A T I O N S Applications of AI Introduction to Artificial Intelligence 7 / 40 Healthcare Diagnosis, imaging, drug discovery Education Adaptive tutoring & grading Finance Fraud detection, trading Transportation Self - driving, route planning Manufacturing Robotics, predictive maintenance Agriculture Crop & yield monitoring Entertainment Recommendation & generation M O D U L E 1 · F O U N D A T I O N S AI vs Natural Intelligence Introduction to Artificial Intelligence 8 / 40 Aspect Artificial Intelligence Natural (Human) Intelligence Learning ability From large datasets; fast on narrow tasks From few examples; broad, transferable Creativity Recombines patterns in training data Original, imaginative, intuitive Adaptability Limited to its training distribution Flexible across novel situations Emotions Simulated, not genuinely felt Genuine emotions and empathy Decision making Consistent, data - driven, tireless Context - rich, value - and ethics - aware Speed & scale Extremely fast, massively parallel Slower, but deeply contextual M O D U L E 1 · C A S E S T U D I E S AI in Practice: Case Studies Introduction to Artificial Intelligence 9 / 40 Autonomous vehicles fuse perception, search - based planning and learned control in real time. ChatGPT Generative AI built on the decoder - only Transformer (GPT) architecture; predicts the next token using self - attention, then is fine - tuned with human feedback (RLHF). Autonomous Vehicles Combine computer vision, sensor fusion and path - search planning to perceive surroundings and navigate safely. Medical Diagnosis Decision - support systems analyse scans and records; human+AI teams improve diagnostic accuracy. Recommendation Systems Netflix and Amazon use collaborative filtering and ranking to personalise content for hundreds of millions of users. M O D U L E 1 · R E V I S I O N Summary & Practice Questions Introduction to Artificial Intelligence 10 / 40 Key Takeaways • AI builds systems that reason, learn, perceive and act • Rational - agent view guides modern AI • Six characteristics; five technique families • Applications span every major industry • AI is narrow; human intelligence is general Viva / Interview Questions • Differentiate weak AI from strong AI. • What is the Turing Test? State one limitation. • Define a rational agent with an example. Multiple - Choice Questions Q1. The term 'Artificial Intelligence' was coined by: (a) Alan Turing (b) John McCarthy (c) Marvin Minsky (d) Herbert Simon Answer: (b) John McCarthy, 1956 Q2. Which is NOT a characteristic of AI? (a) Learning (b) Reasoning (c) Photosynthesis (d) Perception Answer: (c) Photosynthesis M O D U L E 2 · P R O B L E M S & S P A C E S Understanding AI Problems Introduction to Artificial Intelligence 11 / 40 LE AR N ING O B JEC TI VES Define an AI problem · formulate states, operators and goals · represent and search a state space · understand production sys tem s. • Definition • An AI problem is a task defined by an initial situation, a goal, and the actions that move between them. • Characteristics • Well - defined states & goals · clear operators · measurable path cost · often very large search spaces. Types of AI Problems 1 Ignorable Solution steps can be ignored — e.g. theorem proving 2 Recoverable Steps can be undone — e.g. 8 - puzzle 3 Irrecoverable Steps cannot be undone — e.g. chess 4 Toy vs Real Puzzles for teaching vs real - world tasks M O D U L E 2 · P R O B L E M S & S P A C E S Problem Formulation in AI Introduction to Artificial Intelligence 12 / 40 Initial State Where the agent starts Operators / Actions Legal moves that change state Goal Test Checks if the goal is reached Path Cost Sum of step costs along a path Example — Route Finding (Arad → Bucharest) A S F R P B Initial = Arad (A) · Goal = Bucharest (B) · green path is one solution M O D U L E 2 · P R O B L E M S & S P A C E S Problem Space Concept Introduction to Artificial Intelligence 13 / 40 • Definition • The problem space is the set of all states reachable from the initial state by applying operators — the universe the agent searches. • Components • States · operators · initial state · goal state(s) · path cost. • Applications • Puzzles, planning, theorem proving, game playing and route finding. Problem - Space Tree S n1 n2 n3 G S = start · branches = operators · G = goal M O D U L E 2 · P R O B L E M S & S P A C E S State Space Representation Introduction to Artificial Intelligence 14 / 40 • Definition • A state space is a graph whose nodes are states and whose edges are state transitions produced by operators. • State transitions • Each operator maps one state to a successor state. • Represented as • A directed graph (cycles allowed) — unlike a tree, states can be revisited. State - Space Graph A B C D E F Start = A · Goal = F · arrows show legal transitions M O D U L E 2 · P R O B L E M S & S P A C E S State Space Search Introduction to Artificial Intelligence 15 / 40 • Search process • Start at the initial state, expand successors, test for the goal, repeat until found. • Search tree vs graph • The tree records explored paths; the graph view merges repeated states. • Solution path • The sequence of operators from root to the goal node. Search Tree (solution path in green) S A B C D E G H Solution: S → B → G M O D U L E 2 · R E A S O N I N G S Y S T E M S Production Systems Introduction to Artificial Intelligence 16 / 40 • Definition • A production system solves problems using a set of IF – THEN production rules plus a control strategy. • Components • Global database (working memory) · production rules · control / inference engine. • Rule - based • Each rule fires when its condition matches the current state. Production - System Architecture Working Memory Rule Base (Productions) Control / Inference Engine Match → Resolve conflicts → Fire → Update memory M O D U L E 2 · R E A S O N I N G S Y S T E M S Types of Production Systems Introduction to Artificial Intelligence 17 / 40 Monotonic Applying a rule never prevents another rule that was applicable from being applied later. Example: Most theorem proving Non - Monotonic Firing a rule can invalidate rules that were previously applicable. Example: Real - time planning Partially Commutative Order of a permitted rule set affects the path but not the final state. Example: 8 - puzzle, Water Jug Commutative Order of rules affects neither the state reached nor applicability. Example: Database updates M O D U L E 2 · R E A S O N I N G S Y S T E M S Expert Systems & Production Rules Introduction to Artificial Intelligence 18 / 40 IF – THEN Rule (example) IF fever AND cough AND body - ache THEN diagnosis = influenza (confidence 0.8) • Knowledge Base • Domain facts and IF – THEN production rules. • Inference Engine • Applies rules via forward or backward chaining. • Working Memory + UI • Holds case facts; explains its reasoning to the user. Real - World Expert Systems 1 MYCIN Diagnosed bacterial infections (medicine) 2 DENDRAL Inferred molecular structure (chemistry) 3 XCON / R1 Configured computer systems (DEC) 4 PROSPECTOR Mineral exploration & geology 5 CaDet Early cancer detection support M O D U L E 2 · C A S E S T U D Y Case Study: The Water Jug Problem Introduction to Artificial Intelligence 19 / 40 Problem Statement Given a 4 - litre jug and a 3 - litre jug (no markings), measure exactly 2 litres. Operations: fill, empty, or pour between jugs. State = (x, y) where x ≤ 4, y ≤ 3. Solution Strategy • Initial (0,0) → goal any state with x = 2 • Apply operators; search the state space (BFS/DFS) • Detect repeated states to avoid cycles • Solution = sequence of fill/empty/pour moves State - Space Solution Path (0,0) (4,0) (1,3) (1,0) (0,1) (4,1) (2,3) Goal reached at (2,3): the 4 - L jug holds exactly 2 litres M O D U L E 2 · R E V I S I O N Summary & Practice Questions Introduction to Artificial Intelligence 20 / 40 Key Takeaways • A problem = initial state, operators, goal test, path cost • Problem space vs search tree vs state - space graph • Production systems run a match – fire – update cycle • Expert systems apply IF – THEN rules over a knowledge base • Water Jug models problems as (x, y) state tuples Viva / Interview Questions • Differentiate a state space from a search tree. • List the components of a production system. • What is forward vs backward chaining? MCQs & Numerical Practice Q3. A state space is best modelled as a: (a) Queue (b) Directed graph (c) Hash table (d) Heap Answer: (b) Directed graph Q4. The inference engine in a production system: (a) Stores rules (b) Matches & fires rules (c) Renders UI (d) Compiles code Answer: (b) Matches & fires rules Numerical: Using a 5 - L and 3 - L jug, measure exactly 4 L. List every (x,y) state on your path.