Hfrxdjt

I'm working on a binary classification in Keras with a Tensorflow backend. No matter how much I tweak, I can't seem to get my model past a val_acc of 67%. Is there something I'm missing, or is this just simply as accurate as I can get with my data?

Link to the data I am using

My Code

Load and Balance dataset to a 1:1, and create validation data.

from sklearn.utils import resample

raw_data = pd.read_csv('Data.csv')

df_majority = raw_data[raw_data['RESULT']==0].iloc[1:-2,0:3].dropna()
df_minority = raw_data[raw_data['RESULT']==1].iloc[1:-2,0:3].dropna()

print(raw_data['RESULT'].value_counts())


df_majority_downsampled = resample(df_majority, 
 replace=False,
 n_samples=raw_data['RESULT'].value_counts()[1],
 random_state=123)

# Combine minority class with downsampled majority class
df_downsampled = pd.concat([df_majority_downsampled,df_minority])

# Display new class counts
print(df_downsampled['RESULT'].value_counts())
print(numpy.unique(df_downsampled['RESULT']))

X = df_downsampled.iloc[1:-2,0:2].dropna()
Y = df_downsampled.iloc[1:-2,2:3].dropna()

X, XTest, Y, YTest = train_test_split(X, Y, test_size = 0.3, random_state = 0)

print(YTest['RESULT'].value_counts()) #Just a double check to make

Create Model

def create_model(activation):
 model = Sequential()

 model.add(Dense(128,activation=activation,input_dim=2))
 model.add(BatchNormalization())

 model.add(Dense(64,activation=activation))
 model.add(BatchNormalization())
 model.add(Dropout(0.2))

 model.add(Dense(32,activation=activation))
 model.add(BatchNormalization())
 model.add(Dropout(0.2))

 model.add(Dense(16,activation=activation))
 model.add(BatchNormalization())
 model.add(Dropout(0.2))

 model.add(Dense(8,activation=activation))
 model.add(BatchNormalization())
 model.add(Dropout(0.2))

 model.add(Dense(4,activation=activation))
 model.add(BatchNormalization())
 model.add(Dropout(0.2))

 model.add(Dense(2,activation=activation))
 model.add(BatchNormalization())
 model.add(Dropout(0.2))

 model.add(Dense(1,activation='sigmoid'))
 # load weights
 model.load_weights("weights.best.hdf5")
 model.compile(loss='binary_crossentropy', optimizer=Adam(lr=0.0001),metrics=['accuracy']) 

return model
model = create_model('relu')

filepath ="weights.best.hdf5"
checkpoint = ModelCheckpoint(filepath,monitor='val_acc',verbose=1,save_best_only=True,mode='max')
callbacks_list = [checkpoint]

history = model.fit(X,Y,epochs=2000,batch_size=32, shuffle = True,validation_data = (XTest,YTest), verbose = 0, 
 callbacks=callbacks_list)

Predict and get score

from sklearn.metrics import roc_auc_score

predict = model.predict_classes(X)
print(numpy.unique(predict))

#for index,val in enumerate(predict):
#print("Predicted: %s, actual: %s, for val %s"(val[0],Y.iloc[index].values,X.iloc[index].values))

predict = [val[0] for val in predict]
print("ras score: ",roc_auc_score(Y,predict))


predict = model.predict(numpy.array([0.0235,0.5]).reshape(-1,2))
print(predict[0][0])

Result using current model