AI Development Lesson 6: Neural Networks

Neural networks power modern AI — image recognition, language models, speech. PyTorch is the professional choice for building them.

Concepts

// Neuron: takes inputs, applies weights, bias, activation function
// Layer: collection of neurons
// Forward pass: data flows through layers to make prediction
// Loss: how wrong the prediction is
// Backpropagation: calculate gradients (blame) for each weight
// Gradient descent: update weights to reduce loss
// Epoch: one pass through all training data
// Batch: subset of training data for one update

PyTorch

import torch
import torch.nn as nn

# Define a neural network
class SimpleNet(nn.Module):
    def __init__(self, input_size, hidden_size, output_size):
        super().__init__()
        self.layers = nn.Sequential(
            nn.Linear(input_size, hidden_size),
            nn.ReLU(),
            nn.Dropout(0.3),
            nn.Linear(hidden_size, hidden_size),
            nn.ReLU(),
            nn.Linear(hidden_size, output_size)
        )
    
    def forward(self, x):
        return self.layers(x)

model = SimpleNet(input_size=10, hidden_size=64, output_size=1)

# Training loop
optimizer = torch.optim.Adam(model.parameters(), lr=0.001)
criterion = nn.BCEWithLogitsLoss()  # binary classification

for epoch in range(100):
    model.train()
    optimizer.zero_grad()
    output = model(X_train_tensor)
    loss = criterion(output, y_train_tensor)
    loss.backward()
    optimizer.step()
    
    if epoch % 10 == 0:
        print(f"Epoch {epoch}: loss={loss.item():.4f}")