Predicting the Yields of Field Vegetable Using the Multiple Functional Regression Model

摘要

In this paper, we propose a new statistical method that can analyze the yield data of the field vegetable using a functional regression model with multiple functional covariates and scalar response. From the experimental results, we could see the following results. First, through the descriptive statistical analysis, we can see that CO2 level is positively correlated with yield, while humidity has a strong negative correlation with the yield of vegetable, but temperature has a weak negative correlation with yield. Second, through functional regression analysis of three environmental variables, we confirmed that CO2 and humidity have a very significant effect on the yield of vegetable at a significance level of 5% or less. Third, we can see that the vegetable yields is maximum when the CO2 level is approximately 870 to 900 (ppm), and the vegetable yield is the maximum when the humidity is 93 to 95 (%), but when the temperature is between 16 and 18 (°C), there is no significant difference in vegetable yields.