I have attempted this course multiple times before (in MatLab). I am redoing it python and have found the implementation part FAR more fun and exciting this time.
I want to continue to apply what I learned beyond just the assignments, so I was wondering if I have enough knowledge to do my own regression project with public datasets? If so, how would I go about starting this?
Hi @theedwardchoi , welcome to the DeepLearning.ai community
It’s always great and exciting to work on python.
Yes, you can start with public data sets available on multiple sites. Always refer to the course whenever you are stuck at any calculation make sure you understand the assignment clearly and apply the right diagnostic. You might not get everything right on the first attempt but by correcting errors you will surely get the optimal result and those mistakes help you to understand each step in depth.
As the first ML project, I highly recommend you take the titanic project.
You can sample code from Kaggle and GitHub.
The most important thing in a project is data.
First, try to understand the data very well. Use visualization to help you
Then try to handle missing values and nan data.
Then it’s important to deal with feature engineering.
You can see different Notebooks in Kaggle. It gives you an idea about all of the above points.
Awesome, thank you. I also read some of the resource threads that gave great in depth steps to start projects.
Can I start with knowledge uo to week 3 or should i wait to complete the 3 courses first? Thanks again
@theedwardchoi Learning from different resources is always good to gain knowledge.
If you want to start the project after course 2, you can but make sure the data is related to supervised learning.
But I personally recommend you to complete all 3 courses and then in whichever model you think you love to work find the data set accordingly.
In course 3 there are very interesting models too
Also, you can start the project and course 3 simultaneously if you have enough time.
Hi @theedwardchoi ,
It is amazing that you want to get on an independent project to grow your expertise!
Adding to what @VIVEK_PATEL1 has said, I would like to share one of the most important lessons I learned from Prof. Ng:
Start doing it, get fast to the first version, then iterate.
That has been key on my own development.
Please share how it goes! I can’t wait to learn to see what you achieve!