Deep Learning with High School Math (or Less)
Submitted by Aakash N S (@aakashns) on Wednesday, 28 March 2018
You don’t need a PhD or a master’s degree or even a bachelor’s degree in Math/CS to learn and appy deep learning. In most cases, all you need is some programming experience and a quick revision of some high school math e.g. differentiation and matrix multiplication. I’ll show you how you can get up and running in just few hours, and build state of the art deep learning models that solve problems that you care about, and slowly understand the inner workings over time.
The target audience includes anyone who is enthusiastic about deep learning but finds it really hard to get started, because they’re unsure if they have the right skills and/or resources to get started.
Brief outline of the talk:
1. Prerequisites for learning Deep Learning (Basic Python + Basic Linear Algebra)
2. How to build a state-of-the-art image classifier in 5 minutes (to show it’s really possible)
3. The essence of deep learning (without any math or code)
4. Debunking common myths (e.g. you need a lot of data, you need a huge computing resources etc.)
5. Where to start: brief overview of some online courses, books and blogs using which you can get up and running in a few hours, and slowly understand the inner workings over time.
6. (If time permits) How to build a portfolio to get hired as a data scientist.
For a large part of this talk, I will focus on external resources:
http://www.fast.ai/ https://www.manning.com/books/deep-learning-with-python http://neuralnetworksanddeeplearning.com/chap4.html
- An open mind
I’m a deep learning practitioner and the founder of an AI company specializing in computer vision and facial recognition.
For over 4 years, I made many attempts to learn ML & Deep Learning, always starting with the theory, trying to understand all the math and matrix calculus behind deep learning. Each time, I’d spend several weeks learning the concepts, and then give up realizing that I have no idea how I can use my knowledge to even implement a simple feedfoward neural network. All that changed a few months ago, when I came across a couple of books and courses that took a top-down approach, focusing on getting results first, and understanding the math later. This completely changed my perspective towards Machine Learning, and I’m hoping to share that with this talk.