Mis on kvantila regressioon prognooside jaoks intervallidega?

Kvantil regression is a regression method that allows estimating not only the mean value of the target variable but also various quantiles of its distribution. In the context of interval forecasting (e.g., time series), quantile regression helps obtain interval forecasts that reflect the model's uncertainty.

Instead of predicting a single point value (e.g., the mean), the model is trained to predict, for example, the 5th and 95th percentiles, forming a 90% confidence interval for the forecast.

This is especially useful for tasks where understanding the range of possible values is important, not just the expected value, such as in finance or demand forecasting.

Example of using quantile regression with the statsmodels library in Python:

import statsmodels.api as sm
import numpy as np

# X - features, y - target variable
X = sm.add_constant(X)  # add a constant
percentiles = [0.05, 0.5, 0.95]
models = {}
for q in percentiles:
    mod = sm.QuantReg(y, X)
    res = mod.fit(q=q)
    models[q] = res

# Now predictions can be obtained for different percentiles