1
$\begingroup$

The Sizes of both the true label and predicted label are same still, the training accuracy is 0.0

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
from sklearn.metrics import accuracy_score

Data Preprocessing

train=pd.read_csv(r'C:\Users\yashd\Downloads\Datasets\titanic\train.csv')
train=train.dropna()
y_train=np.array(train['Survived'])
train=train.drop('Survived',axis=1)  #removing the label from the data

train=train.drop('PassengerId',axis=1) #removing irrelevant features from the training data
train=train.drop('Cabin',axis=1)
train=train.drop('Name',axis=1)
train=train.drop('Ticket',axis=1)

train['Sex']=np.where(train['Sex']=='male',1,0) #assigning a value of 1 to male and 0 to female
train['Embarked']=np.where(train['Embarked']=='S',1,np.where(train['Embarked']=='Q',2,3))
train['Fare']=(train['Fare']-train['Fare'].mean())/train['Fare'].var()
train['Age']=(train['Age']-train['Age'].mean())/train['Age'].var()

x_train=np.array(train)
x_train=x_train.T
y_train=y_train.reshape(1,-1)

Neural Network with 2 hidden layers ,128 neuorns in the first hidden layers and 64 in the second hidden layer. The output layers consists of single sigmoid Neuron

class FNN:
    def __init__(self):
        self.W1=None
        self.b1=None
        self.W2=None
        self.b2=None
        self.W3=None
        self.b3=None
    def sigmoid(self,x):
        return 1/(1+np.exp(-x))
    def forward_prop(self,x):
        self.Z1=np.dot(self.W1,x)+self.b1
        self.A1=np.tanh(self.Z1)
        self.Z2=np.dot(self.W2,self.A1)+self.b2
        self.A2=np.tanh(self.Z2)
        self.Z3=np.dot(self.W3,self.A2)+self.b3
        self.A3=self.sigmoid(self.Z3)
        return self.A3
    def back_prop(self,x,y):
        self.forward_prop(x)
        m=x.shape[1]
        self.dZ3=self.A3-y
        self.dW3=np.dot(self.dZ3,self.A2.T)/m
        self.db3=np.sum(self.dZ3,axis=1,keepdims=True)/m
        self.dZ2=np.dot(self.W3.T,self.dZ3)*(1-self.A2**2)
        self.dW2=np.dot(self.dZ2,self.A1.T)/m
        self.db2=np.sum(self.dZ2,axis=1,keepdims=True)/m
        self.dZ1=np.dot(self.W2.T,self.dZ2)*(1-self.A1**2)
        self.dW1=np.dot(self.dZ1,x.T)/m
        self.db1=np.sum(self.dZ1,keepdims=True)/m
    def fit(self,x,y,epochs=100,learning_rate=0.01,plot=True,disp_loss=False):
        np.random.seed(4)
        self.W1=np.random.rand(128,x.shape[0])
        self.b1=np.zeros((128,1))
        self.W2=np.random.randn(64,128)
        self.b2=np.zeros((64,1))
        self.W3=np.random.randn(1,64)
        self.b3=np.zeros((1,1))
        m=x.shape[1]
        loss=[]
        for i in range(epochs):
            self.back_prop(x,y)
            self.W1-=learning_rate*self.dW1
            self.b1-=learning_rate*self.db1
            self.W2-=learning_rate*self.dW2
            self.b2-=learning_rate*self.db2
            self.W3-=learning_rate*self.dW3
            self.b3-=learning_rate*self.db3
            logprobs=y*np.log(self.A3)+(1-y)*np.log(1-self.A3)
            cost=-(np.sum(logprobs))/m
            loss.append(cost)
        e=np.arange(1,epochs+1)
        if plot:
            plt.plot(e,loss)
            plt.show()
        if disp_loss:
            print(loss)
    def predict(self,x):
        y=np.where(self.forward_prop(x)>=0.5,1,0)
        return y

F=FNN()
F.fit(x_train,y_train)
y_pred=F.predict(x_train)
print('Predicted Label:',y_pred)
print('True Label:',y_train)
acc=accuracy_score(y_train,y_pred)
print(acc)

Output Loss Plot Loss Plot

$\endgroup$
2
$\begingroup$

Based on your screenshot, it's quite clear that the accuracy isn't 0.0 since the first two predictions match the true labels. So something must be wrong with how the accuracy is calculated.

If you go to sklearn's documentation, you'll see that accuracy_score requires 1-d arrays while it seems that you are feeding it 2-d arrays. My guess is that right now, it compares the elements of your arrays and checks if they are identical. Because you feed a 2-d array, it checks whether all predictions match, which unless you are perfectly correct, will always yield you 0.0.

Doing the following should fix your issue:

acc=accuracy_score(y_train[0], y_pred[0])
$\endgroup$
1
  • 1
    $\begingroup$ Thanks, this worked! $\endgroup$ – deadweight414 May 7 '20 at 7:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.