An analyst developing a predictive model used a data set of 10,000 observations with 20 predictors and one target variable. To deal with possible overfitting, the analyst randomly divided the data into training (7,500 observations) and validation subsets (2,500 observations). She then tried 10 different models using the training data. Each model was then checked using the validation data. One model was found to clearly outperform the others in predictive accuracy using the validation data. What would you say about this analyst's work? Group of answer choices By using training and validation data sets, the analyst avoided overfitting. The validation data set was too small to be provide a firm conclusion. A third data set (test) data should have been used. The analyst should have tried more than 10 different models