Simple Neural Network Prediction with Theano

This function defines a simple neural network using Theano library for linear transformations and ReLU activation, then predicts the output.

Technology Stack : Theano, NumPy

Code Type : Custom function

Code Difficulty : Intermediate

                
                    
import theano
import numpy as np

def neural_network_predict(input_data):
    # Define a neural network with Theano
    input = theano.tensor.fvector('input')
    hidden = theano.tensor.nnet.relu(input.dot(2) + 1)  # Linear transformation and ReLU activation
    output = hidden.dot(3) + 1  # Linear transformation and bias

    # Compile a function to predict the output
    predict_fn = theano.function([input], output)

    # Use the function to predict the output for a given input
    result = predict_fn(input_data)

    return result

# Code information