Short observation sequence: ['yes', 'no', 'no', 'no', 'no', 'yes', 'yes', 'yes', 'no', 'no'] Filtering - distribution over most recent state: sunny 0.622 rainy 0.078 foggy 0.301 Prediction - distribution over next state: sunny 0.573 rainy 0.168 foggy 0.259 Smoothing - distribution over state at each point in time: time 0 sunny 0.288 rainy 0.475 foggy 0.237 time 1 sunny 0.621 rainy 0.098 foggy 0.281 time 2 sunny 0.686 rainy 0.043 foggy 0.271 time 3 sunny 0.622 rainy 0.045 foggy 0.333 time 4 sunny 0.412 rainy 0.115 foggy 0.473 time 5 sunny 0.086 rainy 0.619 foggy 0.295 time 6 sunny 0.047 rainy 0.774 foggy 0.179 time 7 sunny 0.107 rainy 0.706 foggy 0.186 time 8 sunny 0.541 rainy 0.144 foggy 0.315 time 9 sunny 0.622 rainy 0.078 foggy 0.301 Viterbi - predicted state sequence: ['sunny', 'sunny', 'sunny', 'sunny', 'sunny', 'rainy', 'rainy', 'rainy', 'sunny', 'sunny'] Viterbi - actual state sequence: ['rainy', 'foggy', 'sunny', 'sunny', 'sunny', 'foggy', 'foggy', 'rainy', 'foggy', 'sunny'] The accuracy of your viterbi classifier on the short data set is 0.5 The accuracy of your viterbi classifier on the entire data set is 0.668