Particle filtering

Particle filtering is a method used in probability and robotics to estimate an unknown state, like a position, over time. It works by representing possible states with many random samples called particles. Each particle has a probability or weight based on how well it fits the observed data. As new information comes in, particles are updated and resampled to focus on the most likely states. This approach enables systems to track changing conditions accurately, even with noisy or incomplete data, by effectively approximating the true probability distribution of the state through these particles.