Introduction
Stock prediction allow you to visualize any stock's data in graph. Furthermore,
it will predict stock's data for next day.
You are able to manage stock by adding new stocks or remove it.
Here is a demo of video 👇.
Motivation & Purpose
Combining machine learning with financial stock and then make
prediction for stock. This could help stock trader to make decision
whether they should buy or sell.
More
The model for stock prediction is an variant of
DecisionTreeRegressor.
In this project it use
RandomForestRegressor
as model for training and making prediction.
This small project use Docker
to run several services to achieve scalability. And use Streamlit
to build frontend web app.
The service allow user to add or remove stock and react to user's change in real-time.
In addition it is able to learn stock data and make prediction everyday.
All stocks' data are from Yahoo Finance.
Frameworks were used to build this app:
🧰 Scikit-Learn : For building model, learning data and making prediction
🧰 Streamlit : Frontend web app
🧰 Docker : Containerized backend services
Note
This small project build and test on local machine so far. Therefore there is
no live service available. Due to no budget for renting a server in order to
run docker services, I have no intention to put it in public. However it's
source code can be download from GitHub.
Website