- Jaron Ritter
- C.J. Moore
- Liam Andrade
- Steven Senger
- Create a venv
- activate the env
- pip install ipykernel
- run
ipython kernel install --name "nameYouWantForKernel" --user - Verfiy kernel has been set up in jupter lab. Run -
jupyter lab - Install dependencies using the requirements.txt. Run -
pip install -r pathToReqFile/requirements.txt
- Classify images of blood cells as cancerous or not.
- Neural Net
- Random Forest
- If you take a look at the dependencies, there is a libary called plotly. It is a JS based interactive data vis library. I will be using it and I would recommend you look at it too (it's pretty cool)
- If you add a dependency please add it to the requirements.txt
- Please create a new branch off main to work on
- Please keep changes smallish so they're not hard to review
- Have other reivew your code and verify they review it before merging to main
- If you add new packages into it, please update the requirements.txt
- If you need to import a library that had ml algorithms in it, ping me (Jaron) because they need approval by Dr. Saquer
- Determine correct preprocessing steps.
- Should we gray scale or find a way to use all three channels of color?
- Should we save the processed images? (There are a lot of them)
- Feature Selection? (We shouldn’t need it)
- Preprocess data and time it. Does it happen in a reasonable time?
- Test the models (NN vs Random Forest), and select the best
- Do not throw away the results of the losing model. We need it to compare in the write up
- This step should include hyperparameter tuning if applicable
- It says we should also repeat this for validity
- Interpret the results.
- Can we do this, and if so, what does it mean?
- Data Visualization
- Confusion matrix, in our case false negatives will be very bad
- Scatter Plots, this might lead us to try a clustering algorithm
- Note the issues you run into so we can talk about it in the write up
- If you have ideas on how to improve the results or future plans for this project, note it. We will need it for the write up.
- Write the write up (This is going to take the longest)
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent