이산 시간 확률 과정을 이용한 질병 감염의 예측

Prediction of Disease Infection using Discrete-time Probability Process

Tuberculosis is one of the leading causes of mortality and morbidity worldwide. According to the latest report of the National Tuberculosis Association, It is appeared as 5 patients and 1 Mortal per 100,000 people in the United States. Whereas Korea is unfortunately recorded as 90 patients and 10 Mortal per 100,000 people overwhelmingly higher than other OECD member countries. In recent years the incidence of tuberculosis is increasing in 20s people specially. In 19th century tuberculosis has been a lot of pain and gave mankind so called 'the white plague'. In this paper discrete-time Markov chain random process is applied in a representative way for predicting tuberculosis infection. For our experimental process, tuberculosis infection data are used for predicting patient using a time-dependent model. By comparison of the actual value and prediction, the feasibility of predictive modeling infection is examined according to time proposed. As the result the average predicted probability was compared whether to increase the prediction rates.