Join Our 5-Week ML/AI Engineer Interview Bootcamp 🚀 led by ML Tech Leads at FAANGs
Back to Questions
175. Multi layer perceptron inference
easy
Generalsenior
Implement the inference (forward pass) of a simple Multi-Layer Perceptron (MLP) so you can turn an input vector into an output prediction. You’ll compute two linear layers with a ReLU activation in between, using the standard MLP formula:
Requirements
Implement the function
python
Rules:
- Compute the first linear layer: ( z_1 = W_1 x + b_1 ).
- Apply ReLU elementwise: ( h = \max(0, z_1) ).
- Compute the output layer: ( \hat{y} = W_2 h + b_2 ).
- Return the output as a NumPy array.
- Use only NumPy and Python built-in libraries (no deep learning frameworks).
Example
python
Output:
python
Input Signature
| Argument | Type |
|---|---|
| x | np.ndarray |
| W1 | np.ndarray |
| W2 | np.ndarray |
| b1 | np.ndarray |
| b2 | np.ndarray |
Output Signature
| Return Name | Type |
|---|---|
| value | np.ndarray |
Constraints
Return NumPy array
Only NumPy + Python built-ins
Use ReLU between two linear layers
Hint 1
Use np.dot(W, x) for matrix-vector multiplication.
Hint 2
For ReLU, use np.maximum(0, z1) elementwise.
Hint 3
Double-check dimensions: W1 shape (d_hidden, d_in), W2 shape (d_out, d_hidden).
Roles
ML Engineer
AI Engineer
Companies
GeneralLevels
senior
entry
Tags
forward-pass
matrix-vector-multiplication
ReLU
MLP
45 people are solving this problem
Python EditorLn 1, Col 1