NBA Prediction System Instruction
- install Armadillo library, pack included in source code floder.
- install MLPACK library, pack included in source code floder.
- if just testing system, click the index.php file.
- if want to check back-end code using command: g++ -std=c++11 -Wall -lmlpack -O1 -larmadillo -o
Code file usage:
- data_formation.cpp -> formation_data label
- feature_file_processing.cpp -> feature_data date
- norm_processing.cpp -> stddev, normalise, mean data
hmm: 4. state_processing.cpp -> standardization, state data(obeservation sequence, label sequence) 5. hmm.cpp -> hmm model, hmm prediction data 6. simulation_recover.cpp -> sim_re data, performance data, pca data
win rate: 7. cal_overall_average.cpp -> overal average running 8. cal_vh_average.cpp -> host / visitor average running 9. cal_lastN_average.cpp -> last N games average running
- team_data_connection.cpp -> data set each team 11 schedule_formation.cpp -> schedule_formation data
- training_test_set_processing.cpp -> training/test data set
- linear_regression.cpp -> prediction model and result
NBA system CODE
- searchGame.cpp -> return games scheduel on search date
- linearRegressionPredict-> return preidction according to date and team example "20141225LALCHI"