Journey of 100 Days Of ML Code Challenge
I have completed my 100 days of ML code challenge #100DaysofMLCodesuccessfully on 31st Oct 2018.
I started my journey as a newcomer to this exciting world of ML. I started with the basics, learned mathematics concepts via khan academy and various other YouTube channels also understand history and concepts of AI/ ML on edX — AI micro masters, Microsoft Professional Program for Artificial Intelligence help to learn ML, deep learning and to build models using Microsoft Azure Machine Learning studio.
I learned Python language via DataCamp also did data visualization, data prediction etc on the datasets available in the various tracks.
Intro to ML course from udacity help me to understand and implement few important Ml algorithms, identify and evaluate features, assess its performance, also tried them on mini projects available at the end of each section, Also use all the techniques learned on ENRON email dataset as a final project.
Learned basics of deep learning neural networks, NLP, and usage of famous libraries like TensorFlow from DataCamp and YouTube Channel sentdex
Tip: Anyone who wants to understand NLP concepts should definitely check out this channel
As a project, I made an analyzer to identify top trending topics of machine learning basic of Neural Information Processing Systems (NIPS) NIPS research papers. I used NLP to find topics.
As I’m from the Android background I used TensorFlow lite (mobile-friendly version of TensorFlow) and worked on few use cases, (still under development) while learning and implementing I found out, some issues of TensorFlow Codelabs documentation, fixed identified bugs of Google TensorFlow mobile git repo (pull request is under review). You can download/view bug-free Codelabs sample from my GitHub account here
I want to thank Data science community for all the support and motivation and of course, Siraj Raval for this wonderful challenge though the challenge was completed, my ML journey just started.