Narrow Artificial Intelligence with Machine Learning for Real-Time Estimation of a Mobile Agent’s Location Using Hidden Markov Models

Figure 5

Average distance for each game. Each figure represents a different run of the experiment. The green line represents an AI with no machine learning and which used a uniformly distributed transition matrix. The black line represents the same AI but this time one that used a transition matrix built using game experience, as Hladky and Bulitko did. Finally, the red line was obtained by playing against our AI with adaptive memory.