''' Following is an example function of k-fold cross validation. - In this code, you are given the dataset, x_train and y_train. And you try to find the best model with k-fold cross validation. - There are 10 blanks (underlined) for you to fill out. - The goal is not to write out the perfect code, but to understand how k-fold cross validation is actually implemented by studying an example. - Please also note that this is mostly pseudo-code although much of this code is probably runnable code :) ''' def k_fold_cross_validation(x_train, y_train) # x_train and y_train contains all the training data # initial best_model and best_score best_model # variable unassigned at present best_score = 0 # variable initialized to 0 for each model: num_folds = 3 # the number of folds subset_size = len(x_train) / _________ # subset_size = size of each fold of data test_model = sklearn.svm.SVC() or other models # SVC stands for Support Vector Classification. It is a built-in class in # sklearn. We can initialize the model we want to test in this loop correct = 0 # keeps the number of correct predictions for i in range(________): # In k-fold cross validation, we split the dataset into k folds. We use one # fold of data to test the model, and use the rest of the data to train the # model. The following 4 lines of code shows how we extract particular fold # of data from the whole data set. For example, if we want to use the i-th # fold of data to test, given the subset_size, the index of i-th fold data # is from i * subset_size to (i + 1) * subset_size training_this_round_xVal = x_train[:i * subset_size] + x_train[(i + 1) * subset_size:] training_this_round_yVal = y_train[:i * subset_size] + y_train[(i + 1) * subset_size:] testing_this_round_xVal = x_train[________:][:subset_size] testing_this_round_yVal = y_train[________:][:subset_size] # use training data to train the model test_model.fit(________,________) # predict using trained model y_predict = test_model.predict(________) # sum up the correct predicted data for i in range(0, len(y_predict)): if (y_predict[i] == ________): # prediction == actual? correct += 1 # save the best score and best model so far if (float(correct) / len(y_train) > best_score): # If current score better... best_score = float(________) / len(y_train) # ... save the better score best_model = ________ # ... and save current model that resulted in best score return best_score, best_model