Read on Twitter
ARIMA models are essential in Time Series forecasting.
You can add multiple components to make them fit your particular data:
go from a basic AR model to a complex SARIMAX model! 🧵 👇
You can add multiple components to make them fit your particular data:
go from a basic AR model to a complex SARIMAX model! 🧵 👇
🔴 S (Seasonal):
• Represents recurring patterns or variations at fixed intervals in time series data.
• When to consider: when there are predictable, repetitive cycles, such as monthly or yearly patterns.
• Represents recurring patterns or variations at fixed intervals in time series data.
• When to consider: when there are predictable, repetitive cycles, such as monthly or yearly patterns.
🟢 AR (Auto-Regressive):
• Reflects the relationship between the current observation and its past values at lag intervals.
• When to consider: when there's a correlation between the current and past observations, indicating temporal dependence.
• Reflects the relationship between the current observation and its past values at lag intervals.
• When to consider: when there's a correlation between the current and past observations, indicating temporal dependence.
🟡 I (Integrated):
• Represents the number of differences needed to make a time series stationary.
• When to consider: to achieve stationarity, especially in the presence of trends or seasonality.
• Represents the number of differences needed to make a time series stationary.
• When to consider: to achieve stationarity, especially in the presence of trends or seasonality.
🔵 MA (Moving Average):
• Describes the relationship between the current observation and residual errors from a moving average model applied to lagged observations.
• When to consider: when there is evidence of residual correlations after differencing.
• Describes the relationship between the current observation and residual errors from a moving average model applied to lagged observations.
• When to consider: when there is evidence of residual correlations after differencing.
🟣 X (eXogenous):
• Represents additional variables external to the time series that can influence it.
• When to consider: when there are external factors impacting the time series but not inherently part of it.
• Represents additional variables external to the time series that can influence it.
• When to consider: when there are external factors impacting the time series but not inherently part of it.
⚠️ But be careful!
Don't use components that you don't need or you may overfit your data!
Don't use components that you don't need or you may overfit your data!
If you liked this you must definitely check yesterday's thread 👇
https://twitter.com/1531575857705275392/status/1734120967193190775
Join more than 4k 💊MLPills subscribers and enjoy free Machine Learning and Data Science content straight to your email every Thursday!
mlpills.dev/subscribe/
mlpills.dev/subscribe/
You should also join our newsletter, DSBoost🚀
Every week we share:
🔹Interviews
🔹Podcast notes
🔹Learning resources
🔹Interesting collections of content
Subscribe for free👇👇
dsboost.dev
Every week we share:
🔹Interviews
🔹Podcast notes
🔹Learning resources
🔹Interesting collections of content
Subscribe for free👇👇
dsboost.dev
• • •
Missing some Tweet in this thread? You can try to
force a refresh
