def cross_validation_scan(): # v1_train = scan("res/x_validation/v1.train") # v1_test = scan("res/x_validation/v1.test") # v2_train = scan("res/x_validation/v2.train") # v2_test = scan("res/x_validation/v2.test") # v3_train = scan("res/x_validation/v3.train") # v3_test = scan("res/x_validation/v3.test") # v4_train = scan("res/x_validation/v4.train") # v4_test = scan("res/x_validation/v4.test") # v5_train = scan("res/x_validation/v5.train") # v5_test = scan("res/x_validation/v5.test") # v6_train = scan("res/x_validation/v6.train") # v6_test = scan("res/x_validation/v6.test") v1_train = scan("res/xv/f1.train") v1_test = scan("res/xv/f1.xv") v2_train = scan("res/xv/f2.train") v2_test = scan("res/xv/f2.xv") v3_train = scan("res/xv/f3.train") v3_test = scan("res/xv/f3.xv") v4_train = scan("res/xv/f4.train") v4_test = scan("res/xv/f4.xv") v5_train = scan("res/xv/f5.train") v5_test = scan("res/xv/f5.xv") v6_train = scan("res/xv/f6.train") v6_test = scan("res/xv/f6.xv") return {"v1": [v1_train, v1_test], "v2": [v2_train, v2_test], "v3": [v3_train, v3_test], "v4": [v4_train, v4_test], "v5": [v5_train, v5_test], "v6": [v6_train, v6_test]}
import warnings warnings.simplefilter("ignore", RuntimeWarning) from Scanner import scan from Perceptron import perceptron, test_perceptron from LogRegClass import gradient_descent_logistic_reg, test_log_reg_class from SVM import svm, test_svm from Helper import limit_features, feature_scaling, standardiztion mu = 0.25 epoch = 10 train = scan('res/training1.data') train_data = train['d'] train_y = train['l'] test1 = scan('res/test.data/AHU 13.csv') test1_data = test1['d'] test1_y = test1['l'] test2 = scan('res/test.data/AHU38 1.csv') test2_data = test2['d'] test2_y = test2['l'] test3 = scan('res/test.data/AHU19B 1.csv') test3_data = test3['d'] test3_y = test3['l'] train_data = standardiztion(train_data) test1_data = standardiztion(test1_data) test2_data = standardiztion(test2_data) test3_data = standardiztion(test3_data) train_data = limit_features(train_data, [36, 24, 22, 42, 402, 52, 32, 29, 20, 51])
def cross_validation_scan(): v1_train = scan("res/x_validation/v1.train") v1_test = scan("res/x_validation/v1.test") v2_train = scan("res/x_validation/v2.train") v2_test = scan("res/x_validation/v2.test") v3_train = scan("res/x_validation/v3.train") v3_test = scan("res/x_validation/v3.test") v4_train = scan("res/x_validation/v4.train") v4_test = scan("res/x_validation/v4.test") v5_train = scan("res/x_validation/v5.train") v5_test = scan("res/x_validation/v5.test") v6_train = scan("res/x_validation/v6.train") v6_test = scan("res/x_validation/v6.test") return { "v1": [v1_train, v1_test], "v2": [v2_train, v2_test], "v3": [v3_train, v3_test], "v4": [v4_train, v4_test], "v5": [v5_train, v5_test], "v6": [v6_train, v6_test] }
import warnings warnings.simplefilter("ignore", RuntimeWarning) from Scanner import scan from LogRegClass import gradient_descent_logistic_reg, test_weight epochs = 30 sigma = 32 #{'d': training_data, 'l': y_vals} train = scan("res/a5a.train") test = scan("res/a5a.test") # get the training info train_data = train['d'] train_labels = train['l'] # get the test info test_data = test['d'] test_labels = test['l'] w = gradient_descent_logistic_reg(train_data, train_labels, epochs, sigma) # parse w weights = w['w'] log_likelihood = w['o'] # run the test c = test_weight(test_data, test_labels, weights) # get the accuracy accuracy = c["correct"] / (c["correct"] + c["wrong"])
import warnings warnings.simplefilter("ignore", RuntimeWarning) from Scanner import scan from Winnow import winnow, test_winnow from Helper import limit_features, feature_scaling, standardiztion r = 2 epoch = 1 data = scan('res/training1.data') test_data = data['d'] test_y = data['l'] test_data = standardiztion(test_data) test_data = limit_features(test_data, [36, 24, 22, 42, 402, 52, 32, 29, 20, 51]) print("WINNOW") weight_perceptron = winnow(test_data, test_y, epoch, mu) results_perceptron = test_winnow(test_data, test_y, weight_perceptron) print("POSITIVE: " + str(test_y.count(1))) print("NEGATIVE: " + str(test_y.count(-1))) print("") print("WEIGHT: " + str(weight_perceptron)) # get the accuracy accuracy = results_perceptron["correct"] / (results_perceptron["correct"] +
def scanNow(event): scan(setlist) strVar.set("Scan Complete")