An Empirical Study of Machine Learning Algorithms Applied to Modeling Player Behavior in a "First Person Shooter" Video GameUniversity of Wisconsin--Madison, 2002 - 46 pages |
Expressions et termes fréquents
Accelerate task achieve lower error AdaBoost agent artificial intelligence artificial neural network Backward basic actions began to achieve began to stabilize boosting algorithm boosting and bagging Breiman closest enemy ClosestGoal combat Computer Confusion Matrices CurrentFace CurrentMove data set examples data set sizes data was collected domain Enemies in Sector EnemyDistance EnemyHealth EnemySector ensemble methods Error Rate 45 expert player Face Direction Jump false negatives feature set feature vector finite four tasks genetic algorithms hand coded ID3 algorithm input/output pair lower error rates machine learning algorithms Markov decision process misclassified move forward Naïve Bayes NB ID3 neural network ANN Number Enemies number of enemies output partial decision tree person shooter Player Health player modeling posterior probabilities precondition produced by ID3 Pseudo Code rule set sample Set Error Rate shown Soldier of Fortune supervised learning Test Set Error thesis training data training examples tree as produced tune set weights