Neural network formulations notebook

[jalving]

This PR adds a Jupyter notebook that explores different neural network formulations in OMLT. Mainly, it provides the following examples:

• A neural network with sigmoid activation functions formulated using full-space and reduced-space modes
• A neural network with ReLU activation functions formulated with binary variables (BigM) and complementarity conditions
• A neural network with mixed sigmoid and ReLU functions formulated with full-space and complementarity

The PR also adds the file used to generate the data (build_sin_quadratic_csv.py), as well as the data-set itself (sin_quadratic.csv).

[codecov[bot]]

Codecov Report

Merging #29 (b17af7d) into main (c44243e) will not change coverage. The diff coverage is n/a.

@@           Coverage Diff           @@
##             main      #29   +/-   ##
=======================================
  Coverage   88.79%   88.79%           
=======================================
  Files          23       23           
  Lines        1071     1071           
  Branches      160      160           
=======================================
  Hits          951      951           
  Misses         98       98           
  Partials       22       22           

---

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update c44243e...b17af7d. Read the comment docs.

[jalving]

Incorporated Ruth's comments. This should be okay to merge but feel free to review.