Spaces:
Running
Running
Update pages/4_LogisticRegressioin.py
Browse files
pages/4_LogisticRegressioin.py
CHANGED
@@ -1 +1,74 @@
|
|
1 |
-
#LogisticRegression
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
#LogisticRegression
|
2 |
+
|
3 |
+
import streamlit as st
|
4 |
+
import torch
|
5 |
+
import torch.nn as nn
|
6 |
+
import torch.optim as optim
|
7 |
+
import numpy as np
|
8 |
+
import matplotlib.pyplot as plt
|
9 |
+
|
10 |
+
# Define the Logistic Regression Model
|
11 |
+
class LogisticRegressionModel(nn.Module):
|
12 |
+
def __init__(self):
|
13 |
+
super(LogisticRegressionModel, self).__init__()
|
14 |
+
self.linear = nn.Linear(1, 1)
|
15 |
+
|
16 |
+
def forward(self, x):
|
17 |
+
return torch.sigmoid(self.linear(x))
|
18 |
+
|
19 |
+
# Generate synthetic data
|
20 |
+
np.random.seed(0)
|
21 |
+
torch.manual_seed(0)
|
22 |
+
n_samples = 100
|
23 |
+
X = np.random.rand(n_samples, 1) * 10 # Random hours between 0 and 10
|
24 |
+
y = (X > 5).astype(int).flatten() # Pass if study hours > 5, otherwise fail
|
25 |
+
|
26 |
+
# Convert to torch tensors
|
27 |
+
X_tensor = torch.tensor(X, dtype=torch.float32)
|
28 |
+
y_tensor = torch.tensor(y, dtype=torch.float32)
|
29 |
+
|
30 |
+
# Streamlit interface
|
31 |
+
st.title('Logistic Regression with PyTorch')
|
32 |
+
|
33 |
+
# User inputs
|
34 |
+
num_epochs = st.number_input('Number of Epochs', min_value=100, max_value=5000, step=100, value=1000)
|
35 |
+
learning_rate = st.number_input('Learning Rate', min_value=0.0001, max_value=0.1, step=0.0001, format="%.4f", value=0.01)
|
36 |
+
test_hours = st.text_input('Test Study Hours (comma separated)', '4.0, 6.0, 9.0')
|
37 |
+
|
38 |
+
# Initialize the model
|
39 |
+
model = LogisticRegressionModel()
|
40 |
+
|
41 |
+
# Binary Cross Entropy Loss
|
42 |
+
criterion = nn.BCELoss()
|
43 |
+
|
44 |
+
# Stochastic Gradient Descent Optimizer
|
45 |
+
optimizer = optim.SGD(model.parameters(), lr=learning_rate)
|
46 |
+
|
47 |
+
# Training the model
|
48 |
+
loss_values = []
|
49 |
+
for epoch in range(num_epochs):
|
50 |
+
outputs = model(X_tensor)
|
51 |
+
loss = criterion(outputs, y_tensor.view(-1, 1))
|
52 |
+
optimizer.zero_grad()
|
53 |
+
loss.backward()
|
54 |
+
optimizer.step()
|
55 |
+
loss_values.append(loss.item())
|
56 |
+
|
57 |
+
# Plot the loss curve
|
58 |
+
fig, ax = plt.subplots()
|
59 |
+
ax.plot(range(num_epochs), loss_values)
|
60 |
+
ax.set_xlabel('Epoch')
|
61 |
+
ax.set_ylabel('Loss')
|
62 |
+
ax.set_title('Loss Curve')
|
63 |
+
st.pyplot(fig)
|
64 |
+
|
65 |
+
# Evaluation
|
66 |
+
model.eval()
|
67 |
+
test_hours = [float(hour.strip()) for hour in test_hours.split(',')]
|
68 |
+
test_tensor = torch.tensor(test_hours, dtype=torch.float32).view(-1, 1)
|
69 |
+
predictions = model(test_tensor).detach().numpy()
|
70 |
+
|
71 |
+
# Display predictions
|
72 |
+
st.write('## Predictions')
|
73 |
+
for i, test_hour in enumerate(test_hours):
|
74 |
+
st.write(f"Study hours: {test_hour}, Predicted pass probability: {predictions[i][0]:.4f}")
|