from matplotlib import cm import numpy as np from MachineLearn.Classes import Experiment, DataSet, Data from T4.Perceptron import Layered_perceptron_Logistic COLOR = cm.rainbow(np.linspace(0, 1, 5)) learning_rate = 0.01 epochs = 5000 oExp = Experiment() oDataSet = DataSet() base = np.loadtxt("Datasets/dermatology.data", usecols=range(34), dtype=int, delimiter=",") classes = np.loadtxt("Datasets/dermatology.data", dtype=int, usecols=-1, delimiter=",") for x, y in enumerate(base): oDataSet.add_sample_of_attribute( np.array(list(np.float32(y)) + [classes[x]])) oDataSet.attributes = oDataSet.attributes.astype(float) oDataSet.normalize_data_set() for j in range(20): print(j) oData = Data(len(oDataSet.labelsNames), 31, samples=50)
import numpy as np from matplotlib.lines import Line2D from MachineLearn.Classes import Experiment import matplotlib.pyplot as plt oExp11 = Experiment.load("Objects/EXP01_1_LP_20.gzip".format()) oExp12 = Experiment.load("Objects/EXP01_2_LP_20.gzip".format()) oExp13 = Experiment.load("Objects/EXP01_3_LP_20.gzip".format()) oExp14 = Experiment.load("Objects/EXP01_4_LP_20.gzip".format()) oExp15 = Experiment.load("Objects/EXP01_5_LP_20.gzip".format()) COLORS = ['GREEN', 'RED', 'BLUE'] MARKER = ['o', '^', "*"] base1 = np.loadtxt("Datasets/XOR.txt", delimiter=",") print(oExp11) print() print(oExp12) print() print(oExp13) print() print(oExp14) print() print(oExp15) print() def getBestTrain(exp, name): """Etapa 1: Matriz confusao e grafico para melhor treinamento.""" oDataSet = exp.experimentResults[0]
import numpy as np from matplotlib.lines import Line2D from MachineLearn.Classes import Experiment import matplotlib.pyplot as plt oExp11 = Experiment.load("Objects/EXP02_1_LP_20.gzip".format()) oExp12 = Experiment.load("Objects/EXP02_2_LP_20.gzip".format()) oExp13 = Experiment.load("Objects/EXP02_3_LP_20.gzip".format()) COLORS = ['GREEN', 'RED', 'BLUE'] MARKER = ['o', '^', "*"] base1 = np.loadtxt("Datasets/artifitial1.data", delimiter=",") def imprimir_resultado(oexp, name, oData): oDataSet = oexp.experimentResults[0] print( "EXPERIMENTO " + name + " MELHOR RESULTADO MSE: ", oData.params['MSE'], ) RMSE = [] MSE = [] for i in oDataSet.dataSet: RMSE.append(i.params['RMSE']) MSE.append(i.params['MSE']) MSE = np.array(MSE) RMSE = np.array(RMSE) print("\tRMSE MEDIO ", np.mean(RMSE), "DESVIO", np.std(RMSE)) print("\tMSE MEDIO ", np.mean(MSE), "DESVIO", np.std(MSE))
from sklearn.metrics import accuracy_score, confusion_matrix, mean_squared_error from sklearn.model_selection import KFold from T5.Perceptron import multi_Layered_perceptron_Logistic from T5.Perceptron_r import multi_Layered_perceptron_linear import matplotlib.pyplot as plt COLOR = cm.rainbow(np.linspace(0, 1, 5)) LEARNING_RATE = 0.1 epochs = 200 K_FOLD = 5 GRID = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10] oExp = Experiment() oDataSet = DataSet() base = np.loadtxt("Datasets/artifitial1.data", usecols=range(1), delimiter=",") classes = np.loadtxt("Datasets/artifitial1.data", usecols=-1, delimiter=",") for x, y in enumerate(base): oDataSet.add_sample_of_attribute(np.array(list([np.float32(y)]) + [classes[x]])) oDataSet.attributes = oDataSet.attributes.astype(float) oDataSet.normalize_data_set() oDataSet.labels = np.array([classes]).T for j in range(20): slices = KFold(n_splits=K_FOLD) oData = Data(1, 31, samples=50)
from matplotlib import cm import numpy as np from MachineLearn.Classes import Experiment, DataSet, Data from T1.Perceptron import Perceptron COLOR = cm.rainbow(np.linspace(0, 1, 5)) learning_rate = 0.01 epochs = 5000 oExp = Experiment() oDataSet = DataSet() base = np.loadtxt("Datasets/iris_3.data", usecols=range(4), delimiter=",") classes = np.loadtxt("Datasets/iris_3.data", dtype=object, usecols=-1, delimiter=",") for x, y in enumerate(base): oDataSet.add_sample_of_attribute( np.array(list(np.float32(y)) + [classes[x]])) oDataSet.attributes = oDataSet.attributes.astype(float) oDataSet.normalize_data_set() for j in range(20): print(j) oData = Data(2, 31, samples=50) oData.random_training_test_by_percent([100, 50], 0.8) perc = Perceptron(learning_rate) perc.train(oDataSet.attributes[oData.Training_indexes], oDataSet.labels[oData.Training_indexes], epochs)
def binarizer(mat): result = np.zeros((mat.shape[0], 2)) for i in range(mat.shape[0]): result[i, int(mat[i, 0])] = 1 return result COLOR = cm.rainbow(np.linspace(0, 1, 5)) LEARNING_RATE = 0.1 epochs = 300 K_FOLD = 3 GRID_NEURON = [5, 10, 15, 20] GRID_B = [.25, .5, .75, 1] _OPTIMIZER = SGD(lr=LEARNING_RATE, momentum=0.0, decay=0.0, nesterov=False) oExp = Experiment() oDataSet = DataSet() base = np.loadtxt("Datasets/breast-cancer-wisconsin.data", usecols=range(1, 10), dtype=int, delimiter=",") classes = np.loadtxt("Datasets/breast-cancer-wisconsin.data", dtype=int, usecols=-1, delimiter=",") for x, y in enumerate(base): oDataSet.add_sample_of_attribute( np.array(list(np.float32(y)) + [classes[x]])) oDataSet.attributes = oDataSet.attributes.astype(float)
from matplotlib import cm import numpy as np from MachineLearn.Classes import Experiment, DataSet, Data from T3.Perceptron import Layered_perceptron import matplotlib.pyplot as plt from T4.Perceptron import Layered_perceptron_Logistic COLOR = cm.rainbow(np.linspace(0, 1, 5)) learning_rate = 0.10 epochs = 30000 oExp = Experiment() oDataSet = DataSet() base = np.loadtxt("Datasets/column_3C.dat", usecols=range(6), delimiter=" ") classes = np.loadtxt("Datasets/column_3C.dat", dtype=object, usecols=-1, delimiter=" ") for x, y in enumerate(base): oDataSet.add_sample_of_attribute(np.array(list(np.float32(y)) + [classes[x]])) oDataSet.attributes = oDataSet.attributes.astype(float) oDataSet.normalize_data_set() oExp.add_data_set(oDataSet, description=" Experimento COLUNA 3C LP 20 realizaçoes. com 30000 epocas".format()) for j in range(20): print(j) oData = Data(len(oDataSet.labelsNames), 31, samples=50) oData.random_training_test_by_percent([60, 150, 100], 0.8) perc = Layered_perceptron_Logistic(learning_rate, len(oDataSet.labelsNames)) perc.train(oDataSet.attributes[oData.Training_indexes], oDataSet.labels[oData.Training_indexes], epochs)
from matplotlib import cm import numpy as np import matplotlib.pyplot as plt from MachineLearn.Classes import Experiment, DataSet, Data from T2.Perceptron import Perceptron_Adaline COLOR = cm.rainbow(np.linspace(0, 1, 5)) learning_rate = 0.1 epochs = 5000 oExp = Experiment() oDataSet = DataSet() base = np.loadtxt("Datasets/dt_1.txt", usecols=range(1), delimiter=" ") classes = np.loadtxt("Datasets/dt_1.txt", usecols=-1, delimiter=" ") for x, y in enumerate(base): oDataSet.add_sample_of_attribute( np.array(list([np.float32(y)]) + [classes[x]])) oDataSet.attributes = oDataSet.attributes.astype(float) oDataSet.normalize_data_set() classes = np.array([classes]).T for j in range(20): print(j) oData = Data(2, 31, samples=50) indices = np.arange(oDataSet.attributes.shape[0]) np.random.shuffle(indices) oData.Testing_indexes = indices[int(oDataSet.attributes.shape[0] * 0.85):] oData.Training_indexes = indices[:int(oDataSet.attributes.shape[0] * 0.85)] perc = Perceptron_Adaline(learning_rate)
import numpy as np from matplotlib.lines import Line2D from MachineLearn.Classes import Experiment import matplotlib.pyplot as plt oExp11 = Experiment.load("Objects/EXP01_DT1_20.gzip".format()) COLORS = ['GREEN', 'RED', 'BLUE'] base1 = np.loadtxt("Datasets/dt_1.txt", delimiter=" ") # Etapa 1 def getBestTrain(exp): """Etapa 1: Matriz confusao e grafico para melhor treinamento.""" oDataSet = exp.experimentResults[0] best = 1000000 oBestData = None for oData in oDataSet.dataSet: txAcc = oData.params['MSE'] if txAcc < best: best = txAcc oBestData = oData return oBestData oData11 = getBestTrain(oExp11) oDataSet11 = oExp11.experimentResults[0] print("EXPERIMENTO 1 MELHOR RESULTADO", oData11.params) RMSE = [] MSE = []
from matplotlib import cm import numpy as np from MachineLearn.Classes import Experiment, DataSet, Data from T4.Perceptron import Layered_perceptron_Logistic COLOR = cm.rainbow(np.linspace(0, 1, 5)) learning_rate = 0.01 epochs = 5000 oExp = Experiment() oDataSet = DataSet() base = np.loadtxt("Datasets/iris.data", usecols=[2, 3], delimiter=",") classes = np.loadtxt("Datasets/iris.data", dtype=object, usecols=-1, delimiter=",") for x, y in enumerate(base): oDataSet.add_sample_of_attribute( np.array(list(np.float32(y)) + [classes[x]])) oDataSet.attributes = oDataSet.attributes.astype(float) # oDataSet.normalize_data_set() for j in range(20): print(j) oData = Data(len(oDataSet.labelsNames), 31, samples=50) oData.random_training_test_by_percent([50, 50, 50], 0.8) perc = Layered_perceptron_Logistic(learning_rate, len(oDataSet.labelsNames))
from MachineLearn.Classes import Experiment, DataSet, Data from sklearn.metrics import accuracy_score, confusion_matrix, mean_squared_error from sklearn.model_selection import KFold from T5.Perceptron_r import multi_Layered_perceptron_linear import matplotlib.pyplot as plt COLOR = cm.rainbow(np.linspace(0, 1, 5)) LEARNING_RATE = 0.01 epochs = 200 K_FOLD = 5 GRID = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10] oExp = Experiment() oDataSet = DataSet() base = np.loadtxt("Datasets/pmsm_temperature_data.csv", usecols=[0, 1, 2, 3, 4, 5, 6, 7, 9, 10, 11, 12], skiprows=1, delimiter=",") classes = np.loadtxt("Datasets/pmsm_temperature_data.csv", usecols=[8], skiprows=1, delimiter=",") for x, y in enumerate(base): oDataSet.add_sample_of_attribute( np.array(list(np.float32(y)) + [classes[x]])) oDataSet.attributes = oDataSet.attributes.astype(float) oDataSet.normalize_data_set()
from keras.utils import to_categorical from rbflayer import RBFLayer, InitCentersRandom from sklearn.preprocessing import LabelBinarizer import matplotlib.pyplot as plt from T6_2.kmeans_initializer import InitCentersKMeans COLOR = cm.rainbow(np.linspace(0, 1, 5)) LEARNING_RATE = 0.1 epochs = 300 K_FOLD = 3 GRID_NEURON = [20, 15, 10, 5] GRID_B = [.25, .5, .75, 1] _OPTIMIZER = RMSprop(learning_rate=LEARNING_RATE) oExp = Experiment() oDataSet = DataSet() base = np.loadtxt("Datasets/measurements.csv", usecols=range(7), delimiter=",") classes = np.loadtxt("Datasets/measurements.csv", usecols=-1, delimiter=",") for x, y in enumerate(base): oDataSet.add_sample_of_attribute( np.array(list(np.float32(y)) + [classes[x]])) oDataSet.attributes = oDataSet.attributes.astype(float) oDataSet.normalize_data_set() oDataSet.labels = np.array([classes]).T for j in range(10): slices = KFold(n_splits=K_FOLD, shuffle=True) oData = Data(1, 31, samples=50)
import numpy as np from matplotlib.lines import Line2D from MachineLearn.Classes import Experiment import matplotlib.pyplot as plt oExp11 = Experiment.load("Objects/EXP01_1_PS_20.gzip".format()) oExp12 = Experiment.load("Objects/EXP01_2_PS_20.gzip".format()) oExp13 = Experiment.load("Objects/EXP01_3_PS_20.gzip".format()) oExp2 = Experiment.load("Objects/EXP02_PS_20.gzip".format()) COLORS = ['RED', 'BLUE'] print(oExp11) print() print(oExp12) print() print(oExp13) print() print(oExp2) print() # Etapa 1 def getBestTrain(exp, name): """Etapa 1: Matriz confusao e grafico para melhor treinamento.""" oDataSet = exp.experimentResults[0] best = 0 oBestData = None for oData in oDataSet.dataSet: txAcc = oData.get_metrics()[1, -1] if txAcc > best: