Implement a Simple Neural Network | Dataford Interview Questions

Implement a simple neural network in Python.

Asked in the technical interview stage. Implement a fully connected network with one hidden layer to classify binary examples. Use sigmoid activation, batch gradient descent, and backpropagation.

Function

def train_neural_network(X, y, hidden_size, epochs, learning_rate):

X is an n x d list of numeric feature vectors, and y contains binary labels. Return a list of predicted labels after training. Initialize weights deterministically so results are reproducible.

Constraints

1 <= len(X) <= 1000
1 <= len(X[0]) <= 20
1 <= hidden_size <= 20
1 <= epochs <= 10000
0 < learning_rate <= 1
Every row in X has the same length
Each label in y is either 0 or 1

Function Signature

def train_neural_network(X, y, hidden_size, epochs, learning_rate):

Your question is Implement a Simple Neural Network. Start with the requirements in the Question tab.

Run and submit as often as you like. When you're ready, talk me through your approach or go straight to the code.

Problem