I Built a Trading Bot with ChatGPT
GPT Trader is a trading bot I built with chat GPT and I want to show you
how to use it for live trading. At the end of this video, I invested two
thousand dollars. Will I make money or lose money? Find out at the end!
To improve predictions on popular website Kaggle, we can use XG Boost.
Let's take the code that Chat GPT gave us and paste it in a code window
to see what happens. We have to install Tensorflow but the data isn't
legit so let's ask Chat GPT for a better data source. Suraj suggests
signing up for Alpaca to get real-time data from the web today of
whatever stock we want. We'll use the Scikit Learn library with Keras on
top to train a very simple statistical machine learning model on the data,
but that's not going to be enough. To get better results, we need to go
back to our code and modify the stock prediction code.
We recently signed up for Alpaca and got our API key for the paper API,
which means we're not trading with real money yet. So, we've gone back
to the code and replaced that. Now we have a Python script that's pulling
real data from the web. PPO is a way to teach a computer to make
decisions like a human. We're taking this super Advanced technique,
Markov Decision Process, to frame this process in a way that it's
mathematically possible to define a policy. This means our agent, which
is an AI neural network using Proximal Policy Optimization, will act
according to the financial market environment. Once it's trained, all the
training data will be displayed in a back test.
A Cron job is a scheduled task that runs automatically at specified
intervals on Linux or Unix based systems, perfect for what we wanted.
With Python code, this single function can make a trade
programmatically given the secret and public key. I'm going to deploy
this with my white labeled Firebase starter package on my GitHub, called
gbt Trader, which comes with Firebase and everything integrated. We're
going to deploy it to Virtual with terminal right now. I have deposited
money into my Live account but I'm still waiting on that deposit. Once
the deposit is complete, I'm going to let the bot make a trade and come
back in 24 hours to see how much money it has made.
GPT Trader is a trading bot I built with chat GPT and I want to show you
how to use it for live trading. At the end of this video, I invested two
thousand dollars. Will I make money or lose money? Find out at the end!
To improve predictions on popular website Kaggle, we can use XG Boost.
Let's take the code that Chat GPT gave us and paste it in a code window
to see what happens. We have to install Tensorflow but the data isn't
legit so let's ask Chat GPT for a better data source. Suraj suggests
signing up for Alpaca to get real-time data from the web today of
whatever stock we want. We'll use the Scikit Learn library with Keras on
top to train a very simple statistical machine learning model on the data,
but that's not going to be enough. To get better results, we need to go
back to our code and modify the stock prediction code.
We recently signed up for Alpaca and got our API key for the paper API,
which means we're not trading with real money yet. So, we've gone back
to the code and replaced that. Now we have a Python script that's pulling
real data from the web. PPO is a way to teach a computer to make
decisions like a human. We're taking this super Advanced technique,
Markov Decision Process, to frame this process in a way that it's
mathematically possible to define a policy. This means our agent, which
is an AI neural network using Proximal Policy Optimization, will act
according to the financial market environment. Once it's trained, all the
training data will be displayed in a back test.
A Cron job is a scheduled task that runs automatically at specified
intervals on Linux or Unix based systems, perfect for what we wanted.
With Python code, this single function can make a trade
programmatically given the secret and public key. I'm going to deploy
this with my white labeled Firebase starter package on my GitHub, called
gbt Trader, which comes with Firebase and everything integrated. We're
going to deploy it to Virtual with terminal right now. I have deposited
money into my Live account but I'm still waiting on that deposit. Once
the deposit is complete, I'm going to let the bot make a trade and come
back in 24 hours to see how much money it has made.
