Original article was published on Deep Learning on Medium
(My first CNN)
Get the bigger picture.
For my second-to-last project in my DS bootcamp, I had a few options to pick from. I could have processed words with NLP, make a recommendation system, or make a CNN for image recogntion. I chose the CNN and could not be more pleased. Well… if I had time to do all three, I would be even more pleased.
I used around 6,000 x-rays of people with and without pneumonia to create an image recognition algorithm using Deep learning, Keras, and a CNN. This did not seem as overwhelming as previous projects. It seemed very much like mixing and mastering music. They both require a mentality geared around parameters and tuning parameters for the sake of efficiency.
My readme file, presentation,and code quality and organization have been rapidly improving. At this point, I look back at what used to trouble me in April with a broader sense of troubleshooting and understanding. I find myself more fascinated and less intimidated by code.
As I wrap this second-to-last project up, I am wondering what my final project should be on. I have more range in ability and theme with what I am able to work on for what is called my “Capstone” project. I am another step closer to being a known figure in the Data Science community. With more creativity, time spent learning, and trying new techniques, I am sure that will be able to leave my mark in the tech world!