Interpretation of coefficient of determination.
If computing a causal linear regression model of y = a + bX and resultant r2 is very near zero, then one would be able to conclude that
i. y = a + bX is a good forecasting method.
ii. Y = a + bX is not a good forecasting method.
iii. a multiple linear regression model is a good forecasting method for the data.
iv. a multiple linear regression model is not a good forecasting method for the data