Low-rank matrix approximation is a simple way to compress a matrix while keeping most of its information, often done using Singular Value Decomposition (SVD). Your task is to compute a rank-$k$ approximation $A_k$ of a given matrix $A$ that minimizes reconstruction error.
The Eckart-Young-Mirsky theorem states that the optimal rank-$k$ approximation in the Frobenius norm is given by the truncated SVD:
$$
A_k = \sum_{i=1}^k \sigma_i u_i v_i^T = U_k \Sigma_k V_k^T
$$
where $U_k$ and $V_k$ contain the first $k$ columns of $U$ and rows of $V^T$ (columns of $V$), and $\Sigma_k$ contains the top $k$ singular values.

## Requirements

Implement the function
```python
import numpy as np

def low_rank_approximation(A, k):
    \"""
    Computes the rank-k approximation of matrix A using SVD.

    Args:
        A (np.ndarray): Input matrix of shape (m, n).
        k (int): Target rank for the approximation.

    Returns:
        np.ndarray: The reconstructed rank-k matrix A_k of shape (m, n).
    \"""
    # Your solution below
```

**Rules:**
- Use SVD to approximate $A$ as $A \approx U \Sigma V^\top$, then keep only the top $k$ singular values/vectors.
- Return $A_k$ as a NumPy array.
- Do not use any helper libraries beyond **NumPy** and Python built-ins.

## Example

```python
A = np.array([
    [3.0, 2.0, 2.0],
    [2.0, 3.0, -2.0]
])
k = 1

low_rank_approximation(A, k)
```

Output:
```python
[[2.5 2.5 0. ]
 [2.5 2.5 0. ]]
```

Email Counts by Label

Island Perimeter

Low rank matrix approximation

Solve this easy level ML coding question from General. The premier LeetCode for ML platform to master Python, SQL, and Machine Learning algorithms.

480. Low rank matrix approximation

Examples

Constraints

Similar Questions

Argument	Type
A	np.ndarray
k	int