Store Demand Forecast
As in the previous problem, slightly different models are currently used to forecast demand, which is more complicated than the usual ARIMA. Which, for example? SARIMAX.
This model is very similar to the ARIMA model, except that there is an additional set of autoregressive and moving average components.
The SARIMA model allows data to be distinguished by seasonal frequency as well as other non-seasonal differences. Knowing which options are the best can be made easier with automatic option search frameworks like pmdarina.
You can use SARIMA with statsmodels:
from statsmodels.tsa.statespace.sarimax import SARIMAX
model = SARIMAX(dataset, order=(1, 1, 1), freq="D")
results = model.fit()
The moving average can also be used to predict demand. However, the results we can get using this method can surpass even XGBoost (reduces the error by 32%). But what should we expect from such a simple method?
In any case, your main task in time series prediction is the optimal choice of the model size (its computational performance) and the results it can bring.
Merci pour vos commentaires !
Demandez à l'IA
Demandez à l'IA
Posez n'importe quelle question ou essayez l'une des questions suggérées pour commencer notre discussion
Posez-moi des questions sur ce sujet
Résumer ce chapitre
Afficher des exemples du monde réel
Awesome!
Completion rate improved to 3.85Store Demand Forecast
Glissez pour afficher le menu
As in the previous problem, slightly different models are currently used to forecast demand, which is more complicated than the usual ARIMA. Which, for example? SARIMAX.
This model is very similar to the ARIMA model, except that there is an additional set of autoregressive and moving average components.
The SARIMA model allows data to be distinguished by seasonal frequency as well as other non-seasonal differences. Knowing which options are the best can be made easier with automatic option search frameworks like pmdarina.
You can use SARIMA with statsmodels:
from statsmodels.tsa.statespace.sarimax import SARIMAX
model = SARIMAX(dataset, order=(1, 1, 1), freq="D")
results = model.fit()
The moving average can also be used to predict demand. However, the results we can get using this method can surpass even XGBoost (reduces the error by 32%). But what should we expect from such a simple method?
In any case, your main task in time series prediction is the optimal choice of the model size (its computational performance) and the results it can bring.
Merci pour vos commentaires !
