Beyond Traditional Pathfinding: Focus on Solutions, Not Paths
Explore Hill Climbing, Simulated Annealing, and optimization techniquesLocal search represents a paradigm shift from traditional AI search. Instead of exploring paths through state spaces, local search focuses on iteratively improving solutions by making small, local changes. This approach is ideal for optimization problems where the journey doesn't matter - only the destination.
Interactive introduction to the N-Queens problem and utility function concepts. Place queens on a 4×4 board and watch how utility values change with conflicts!
Explore the classic TSP optimization problem. Understand tour representations, utility functions, and 2-opt local moves through interactive visualizations!
Watch step-by-step hill climbing execution on 4×4 N-Queens. See how neighbor evaluation and utility functions guide the search - or get stuck in local maxima!
Interactive TSP optimization with 2-opt moves. Watch ALL neighbor evaluations and see how hill climbing finds local optima in the tour distance landscape!
Compare city swaps vs 2-opt moves! Visual comparison showing why 2-opt eliminates crossings and finds better tour improvements than simple swaps.
Interactive 2D landscape showing local maxima, plateaus, shoulders, and ridges! Place the hill climbing agent anywhere and watch it navigate different terrain challenges.
Classic textbook visualization! Interactive 1D curve showing global maximum, local maximum, shoulder, and flat plateau. Perfect for understanding hill climbing fundamentals.
Step-by-step N-Queens execution with detailed acceptance probability calculations! Watch temperature control bad move acceptance with P = e^(ΔE/T) formula visualization.
Apply simulated annealing to the Traveling Salesman Problem! Interactive visualization showing tour optimization with temperature-controlled 2-opt moves and acceptance decisions.
Evolution-based optimization with population, selection, crossover, and mutation! Watch solutions evolve over generations with interactive visualization and parameter control.
Interactive comparison of Roulette Wheel, Tournament, and Rank selection methods! Visual demonstrations showing how each method chooses parents with different probabilities.
Comprehensive introduction to local search concepts. Compare classical vs local search, understand state spaces, and learn the fundamentals of optimization algorithms!
Complete step-by-step exercise comparing Hill Climbing vs Simulated Annealing on a 4×4 utility grid. Perfect for exams with detailed solutions and visual grid maps!
Hands-on GA practice with mini timetabling problem. Learn selection, crossover, and mutation through chromosome visualization and fitness calculations!
The simplest local search algorithm. Always moves to the best neighboring state. Fast but can get stuck at local maxima.
Examines all neighbors before choosing the best one. More informed decisions but higher computational cost per step.
Overcomes local optima by running hill climbing multiple times from different starting points. Higher success rate.
Allows occasional "bad" moves to escape local optima. Probability of bad moves decreases over time like cooling metal.
Maintains a memory of recent moves to avoid cycling. Explores more systematically than pure local search.
Population-based approach inspired by evolution. Maintains multiple candidate solutions and combines them.
Place N queens on a chessboard so none attack each other. Perfect for local search since only final configuration matters.
Find shortest tour visiting all cities. Huge state space makes traditional search impractical - local search excels here.
Assign courses to time slots and rooms optimally. Complex constraints and large solution space ideal for local search.
Optimize circuit layout for minimal area and maximum performance. Local search helps fine-tune component placement.
Train neural networks by adjusting weights to minimize error. Gradient descent is a form of local search.
Optimize manufacturing schedules, resource allocation, and supply chain logistics using local search techniques.