https://github.com/komangandika/arima-and-friends
Time Series Forecasting with statistical model such as, AR, MA, ARMA, ARIMA, and SARIMA
https://github.com/komangandika/arima-and-friends
arima forecasting machine-learning price-prediction time-series
Last synced: 9 months ago
JSON representation
Time Series Forecasting with statistical model such as, AR, MA, ARMA, ARIMA, and SARIMA
- Host: GitHub
- URL: https://github.com/komangandika/arima-and-friends
- Owner: KomangAndika
- Created: 2024-06-28T07:58:30.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2024-07-08T08:43:30.000Z (over 1 year ago)
- Last Synced: 2025-01-21T16:26:03.937Z (10 months ago)
- Topics: arima, forecasting, machine-learning, price-prediction, time-series
- Language: Jupyter Notebook
- Homepage:
- Size: 3.47 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# ARIMA-And-Friends
Explore and implement time series forecasting techniques using fundamental statistical models including Autoregressive (AR), Moving Average (MA), Autoregressive Moving Average (ARMA), Autoregressive Integrated Moving Average (ARIMA).
I forecast BTC-USD price
Of course there are some limitation like:
- I am not well-versed in forecasting/time series problem.
- Not enough data(only 350-ish were used in this project).
- This is univariate so only one variable is used which may caused bias.
- Granularity that I am choosing is may not suitable and causing white noise.
- The model that I am using is not as sophisticated like RNN or LSTM.
- There might be seasonality in the data which I can't seem to capture and making my prediction tanking :(
While this might be incomplete, but in the future i want to incorporate more sophisticated method like RNN, LSTM, and XGBoost