Deep Optimal Stopping with Google TensorFlow

It is time to get rich with the deep optimal stopping technology developed with google tensorflow machine learning

Comment 1

Inspiration

Initially, me and my friends were planning to make a Kerr Black Hole photon trajectory simulation to get a pretty image, which would have been a sick project. However, all of my teammates decided to drop out last minute and I was the only one left out of the group to enroll in this hackathon.

I had to come up with something on the spot, and I decided to do this machine learning crypto trade instead, which seemed a lot more manageable.

What it does

  1. Parses the historical data from the cryptocurrency exchange Binance
  2. Uses Google TensorFlow machine learning api to acquire optimal stopping strategy
  3. Automates trading

How I built it

Python with whole bunch of known packages!!! It is amazing how much support the python community has for the language and is making amazing APIs!!!

Challenges I ran into

TensorFlow is written for very specific architectures, so my cpu had difficulty running it. Also, it had collisions with Microsoft c++ redistributable, NVIDIA CUDA drivers, and all sorts of other dlls. I know that TensorFlow is an amazing API but it really was hard to get it compatible. It sort of forced me to use Google Cloud Machine Learning.

Accomplishments that I'm proud of

This is my very first time doing machine learning and parsing data through API, this experience allowed me to figure things out on the spot.

What I learned

Many many APIs and packages! and Google Cloud Computing was exciting!

What's next for BeachHacksCrypto

Anything that involves fun and crazy ideas!

Built With

Share this project:

Updates