15. Performance of Autoregressive Tree Model in Forecasting Cancer Patients

In this paper, our aim is to study the time series data on cancer patients of Punjab, a state of India, by applying the autoregressive tree (ART) model. ART model is an extension of AR model which provides a superior predictive accuracy as compared to AR. In ART model, at first predicted variable and target variable are to be decided on the basis of our own data. And the second step is to form a decision tree for our target variable by transforming the data. This approach of AR model is very much effective for forecasting and a very few research work had been done up to date with ART model. We choose this region for our research purpose as Punjab is considered as a cancer capital of India due to its high increasing rate of cancer-affected patients. In this paper, our goal is to forecast the future trend of cancer patients of Punjab and a case study of cancer incidence is also analyzed and forecast trends of common cancer types among male, female, and both of them. A hypothesis test and calculating our forecast result are also provided in this paper on the basis of mean forecast error and mean square forecast error. The purpose of this study is to aware the people and various organizations regarding the frightened issue of this region and our results could provide appropriate planning regarding eradicate this disease.