homework 2

neural newtork fundamentals

In this homework, we will gain the fundamentals of neural networks, implementing a two layer neural network by hand. An outsized portion of this homework (more than others) is rooted in linear algebra fundamentals, Here are some mathematical properties that will help you in derivations. Review the homework in this pdf file. Remember that reading resources can be found in the syllabus.

data and starter kit

Along with some helpful properties, you will need the utility functions and twitter data at our website. You can also feel free to use the homework template assignment2.py. Please develop in Python without the aid of libraries (e.g., tensorflow, keras, pytorch, jax, etc.) besides numpy, as our autograders will be grading accordingly. If you are more comfortable with notebooks, there are several options:

• Locally on Your Laptop
• Google Cloud Vertex Work with your Google Cloud credits.
• Google Colabs with your Google Account

Document templates can be either Overleaf TeX File or DOCX File. When you’ve compiled/finished writing, download the PDF from Overleaf/Google and upload it to the submission link.

submission instructions

Submit via Gradescope before 5pm, Thursday, January 30. Your artifacts will include:

• Compiled (or exported) PDF into a file called assignment2.pdf
• Data and parameters into a Python pickle file called assignment2.pkl
• All code with included functions in a file called assignment2.py