Modeling Advice and Lessons Learned Working at a Company 🌱

Last updated on March 13, 2021

Make sure to provide context before I start diving into results.
If your models aren’t working as expected, try to open up a new Jupyter notebook, remaking them, and then running them. This gives you a fresh start for your code and lets you reexamine whether or not things are working correctly.
If someone mentions something interesting (a paper, open source project, etc.) in a relevant company slack/chat or in a meeting, then I should look into what they mentioned. It is most likely something relevant/useful
If you are unsure of the purpose of a task/what the task is ask for clarification
If you want to go above and beyond on something ask if that is useful or not before you do it, because it is important to prioritize your time on the most important issues
If you are trying to do exploratory analysis to decide on the design of something, you should do the following:
1. Document your major findings. It is easy to forget what you have done and the most important take aways from your analysis so far, if you’ve done them weeks ago
2. Have a plan of what you want to know, why you want to know it, and how you are going to get the information
It’s important to do apples to apples comparisons, if you change multiple things at once you can’t really isolate the effect you’re looking for.
Prioritize your work so that you can be more efficient and do things in parallel. For example, if you want to run an experiment and also implement some feature in production, write the code for the experiment first so that you can run the experiment online while working on implementing that feature into production.

Notes mentioning this note

There are no notes linking to this note.

Here are all the notes in this garden, along with their links, visualized as a graph.