4. Optimizing Exponential Smoothing: The table below shows the optimal smoothing constants from applying exponential smoothing to data, using automated model selection: Level 1.000 Trend 0.000 Seasonality 0.246 (a) Using a series partitioned into training validation peri- ods and the ets function in R, the optimized smoothing constants are those that maximize (choose one of the fol- lowing): . The likelihood over the training period . The likelihood over the validation period . The likelihood over the training period with a penalty on the number of parameters . The likelihood over the validation period with a penalty on the number of parameters (b) The value of zero that is obtained for the trend smoothing Constant means that (choose one of the following): . There is no trend . The trend is estimated only from the first two points . The trend is updated throughout the data . The trend is statistically insignificant (c) What is the danger of using the optimal smoothing con- stant values