Code: print(“newbie”)
“If you can’t explain it simply, you don’t understand it well enough.”
— Albert Einstein
As a neuroscientist, I am well-versed in writing scientific papers and conveying thoughts using technical jargon. But the real challenge isn’t in writing such papers, the real challenge is in being able to understand the concepts, apply them to real-life problems, and generate solutions.
There is a ton of information out there on programming, data science, and AI but as a beginner, the task of sifting through it is quite daunting. After all, where does one even begin?
So, my aim with starting the publication Deep Thinking is to explain what I’m learning, walk you through my process, and share some problem-solving sets. But before we dive deep, I want to show you what’s in my toolbox and encourage you to build your own.
I began by learning one of the easier programming languages, Python. Much of the open source data from AI (research) institutes are scripted in Python and that makes application easy. Some online courses to get started include Codecademy, Udemy, and Coursera. (Codecademy is free and interactive; it is where I started).
Secondly, I familiarized myself with the following resources:
— GitHub: a.k.a developer heaven. It allows developers to create, make changes to, and collaborate on code. Even users who are not developers like me can access a project code and play around.
— Google Cloud Platform: a cloud computing service that will allow you to run ML code in the cloud using Google’s processors.
— Google Colaboratory: many open source ML scripts are available on GitHub and can be run through GC. For example, I recently played around with DeepDream, a pretrained neural network that enhances patterns in images. Watch Tesla change as the code works its magic… pretty neat, right?
— Jupyter: a web application that allows you to create documents containing code and to visualize data.
— TensorFlow: a ML platform.
— Pandas, NumPy, Keras: Python-specific analyses tools.
Lastly, find a mentor or a friend to talk to, someone to troubleshoot with. Engage with communities like Kaggle, LearnProgramming, and Stack Overflow because problems can only be solved by talking about them.
In subsequent articles and as I progress in my journey, I will walk through how to use these tools and how to apply them to specific data-driven problems.
