Testing self-built roboadviser

in #technology7 years ago

I've built my own Python-based financial predictive model, and now I'm sharing my tests.

Hi buddies. Since the epic-battle between AlphaGo vs Lee Sedol, many great generous AI scientists have offered their software for free, like Google's world-famous Tensorflow and Scikit Learn's variety of algorithms like the decision tree.

I've spent lots of time testing them all, and finally found out that decision tree is the best for financial prediction. I tried many times myself, and I got 8 victories out of 10 attempts. So, I think it's going to be exciting for us to watch together.

DO NOT TAKE THIS AS ADVICE YET. IT'S STILL JUST A TEST RUN.

The test is Australian Dollar vs US Dollar. And I entered a paper trading at the arrow-pointed position. The market can be seen at TradingView. The predicted exit point will hopefully be roughly around the cross point. The 2% may happen but... Fingers crossed.

my_entry

Coin Marketplace

STEEM 0.20
TRX 0.24
JST 0.037
BTC 97001.71
ETH 3368.75
USDT 1.00
SBD 3.07