trainData = pd.read_table('../dataset/dataset1/train.txt',
header=None,
encoding='gb2312',
delim_whitespace=True)
testData = pd.read_table('../dataset/dataset1/test.txt',
header=None,
encoding='gb2312',
delim_whitespace=True)
trainLabel = np.array(trainData.pop(3))
trainData = np.array(trainData)
testLabel = np.array(testData.pop(3))
testData = np.array(testData)
time_start1 = time.time()
clf1 = DecisionTreeClassifier()
clf1.train(trainData, trainLabel)
clf1.predict(testData)
score1 = clf1.accuarcy(testLabel)
time_end1 = time.time()
print("Accuracy of self-DecisionTree: %f" % score1)
print("Runtime of self-DecisionTree:", time_end1 - time_start1)
time_start = time.time()
clf = RandomForestClassifier()
clf.train(trainData, trainLabel)
clf.predict(testData)
score = clf.accuarcy(testLabel)
time_end = time.time()
print("Accuracy of RandomForest: %f" % score)
print("Runtime of RandomForest:", time_end - time_start)
# Splitting data
from sklearn.model_selection import train_test_split
X = data.drop("label", axis=1).values
y = data["label"].values
X_train, X_test, y_train, y_test = train_test_split(X,
y,
test_size=0.2,
random_state=42)
#%%
# My Algorithm
from RandomForest import RandomForestClassifier
model = RandomForestClassifier(nb_trees=200,
max_depth=50,
n_estimators=300,
max_workers=5)
model.fit(X_train, y_train)
print(model.score(X_test, y_test))
joblib.dump(model, "data/my_random_forest.joblib")
#%%
from sklearn.ensemble import RandomForestClassifier as RFC
rf = RFC(max_depth=50,
n_estimators=300,
criterion='entropy',
verbose=0,
max_features="sqrt")
rf.fit(X_train, y_train)
larangan = [
labeltrain[x][y] for x in range(len(labeltrain)) for y in range(0, 3)
if y == 1
]
informasi = [
labeltrain[x][y] for x in range(len(labeltrain)) for y in range(0, 3)
if y == 2
]
trainAnjuran = [trainf[x][1:] + [anjuran[x]] for x in range(len(trainf))]
trainLarangan = [trainf[x][1:] + [larangan[x]] for x in range(len(trainf))]
trainInformasi = [
trainf[x][1:] + [informasi[x]] for x in range(len(trainf))
]
# In[57]:
# ## Build Model & Prediction
modelA = RandomForestClassifier(rf_trees=80, rf_samples=1000)
modelL = RandomForestClassifier(rf_trees=80, rf_samples=1000)
modelI = RandomForestClassifier(rf_trees=80, rf_samples=1000)
modelA.fit(trainAnjuran)
modelL.fit(trainLarangan)
modelI.fit(trainInformasi)
predictA = []
predictL = []
predictI = []
for y in range(0, len(testf)):
predictA.append(modelA.predict(testf[y][1:]))
predictL.append(modelL.predict(testf[y][1:]))
predictI.append(modelI.predict(testf[y][1:]))
prediction = [[predictA[x]] + [predictL[x]] + [predictI[x]]
for x in range(len(predictA))]
hammingloss, index = processor.getHammingLoss(labeltest, prediction)
test.append(fold1)
def split(dtrain, dtest):
train = [dtrain[x][0:-2] for x in range(len(dtrain))]
test = [dtest[x][0:-2] for x in range(len(dtest))]
labeltrain = [dtrain[x][-1] for x in range(len(dtrain))]
labeltest = [dtest[x][-2] for x in range(len(dtest))]
return train, test, labeltrain, labeltest
index = []
performance = []
for x in range(5):
trainf, testf, labeltrain, labeltest = split(train[x], test[x])
model = RandomForestClassifier(rf_trees=80, rf_samples=1000)
trainfix = [
trainf[x][1:] + [labeltrain[x]] for x in range(len(labeltrain))
]
model.fit(trainfix)
prediction = []
for y in range(0, len(testf)):
prediction.append(model.predict(testf[y][1:]))
hammingloss, index = (processor.getHammingLoss(labeltest, prediction))
index.append(index)
performance.append(hammingloss)
# ## Split Data
#train, validation, labeltrain, labelval = train_test_split(extraction, label, test_size=0.30, random_state=42)
#anjuran = [labeltrain[x][y] for x in range(len(labeltrain)) for y in range(0,3) if y==0]
#larangan =[labeltrain[x][y] for x in range(len(labeltrain)) for y in range(0,3) if y==1]
#informasi = [labeltrain[x][y] for x in range(len(labeltrain)) for y in range(0,3) if y==2]