INFO 527 — Neural Networks

Assignment 1: String-to-Vectors (stv_nn)

Objectives

The objectives of this assignment are to:

1. Practice Python programming and use of NumPy arrays.
2. Understand the importance of testing code for correctness and reliability.
3. Implement an Index class that maps string or object sequences to numeric vectors and matrices, a common preprocessing step in neural network training.

---

Files in This Repository

File	Description
notebooks/stv_nn.ipynb	Main notebook containing the full implementation of the Index class.
notebooks/string-to-vector.ipynb	Notebook used to run pytest validation and record test results.
tests/test_nn.py	Provided test file used to validate correctness using pytest.
stv_nn.py	The exported version of the main notebook used for testing.
stv_nn.html	Exported HTML version of the notebook for grading and review.
README.md	This file, describing objectives, setup, and instructions.
requirements.txt	Python dependencies for running tests.

---

Class Overview: Index

The Index class provides a numeric mapping for sequences of objects (typically strings). It supports:

• Creating object-to-index and index-to-object mappings.
• Converting between object sequences and numeric arrays.
• Generating binary (one-hot) encodings and decoding them back to objects.
• Handling padding and unknown objects gracefully.

Implemented Methods:

1. __init__ – Assign indexes to unique vocabulary items starting from a given value.
2. objects_to_indexes – Convert a sequence of objects to a numeric index vector.
3. objects_to_index_matrix – Convert sequences of objects to an indexed matrix with padding.
4. objects_to_binary_vector – Produce a binary (one-hot) vector representation.
5. objects_to_binary_matrix – Produce a binary matrix for multiple sequences.
6. indexes_to_objects – Convert indexes back to their original objects.
7. index_matrix_to_objects – Convert index matrices back to object sequences.
8. binary_vector_to_objects – Retrieve objects from a binary vector.
9. binary_matrix_to_objects – Retrieve objects from a binary matrix.

---

Environment Setup

This assignment reuses the shared virtual environment created for previous coursework.

# Activate the shared environment
source ~/venvs/ml-env/bin/activate

# Navigate to the assignment folder
cd ~/projects/info527-neural-networks-assignment1

# Install dependencies
pip install -r requirements.txt

requirements.txt

numpy
pytest

---

Testing Instructions

1. Run all cells in stv_nn.ipynb to complete the implementation.

2. Export the notebook as:

   • stv_nn.ipynb
   • stv_nn.html
   • stv_nn.py

3. Place the exported .py file in the same directory as test_nn.py.

4. Run pytest to verify correctness:

pytest tests/test_nn.py

Expected Output:

============================= test session starts ==============================
collected 8 items

test_nn.py ........                                                      [100%]

============================== 8 passed in 1.XXs ===============================

---

Submission Files

Submit the following for grading:

• stv_nn.ipynb
• stv_nn.html
• stv_nn.py
• string-to-vector.ipynb (with pytest results)

---

Grading Criteria

Grading is based solely on the correctness of the code written in stv_nn.ipynb.

Method	Description	Marks
__init__	Index initialization	3
objects_to_indexes	Convert objects to index vector	1
objects_to_index_matrix	Convert sequences to indexed matrix	3
objects_to_binary_vector	One-hot vector encoding	2
objects_to_binary_matrix	One-hot matrix encoding	1
indexes_to_objects	Decode index vector	1
index_matrix_to_objects	Decode index matrix	1
binary_vector_to_objects	Decode binary vector	2
binary_matrix_to_objects	Decode binary matrix	1

Total: 15 marks

---

Author

This repository was completed as part of INFO 527: Neural Networks, under the M.S. in Information Science and Machine Learning program (2023–2025).

---