Transformation of the time series:

Univariate model:

- ARIMAx
- Transfer functions
- Exponential smoothing
- Structural unobserved components
- Hodrick-Prescott filter
- Spline smoothing

Multivariate model:

Non-linear model:

- TAR
- ARCH
- GARCH
- Stochastic volatility
- Dynamic Conditional Score
- Generalized Autoregressive Score
- Multiplicative unobserved components

Others :

Genetic model:

Transformation and decomposition:

- Log and Box-Cox
- Detrending
- Deseasonalization
- Autocorrelation
- Cox-Stuart test
- Wavelet transform
- Fourier transform
- STFT
- D.W.T.
- CWT
- Dickey-Fuller Test

Choice of variables and models:

Validation:

Scale-dependent measurement:

Percentage :

In relative terms:

Bayesian model:

Performance index:

Contents

Toggle## Prediction and Forecast/Forecasting

Forecasting and prediction are methods used to estimate future outcomes, but they differ in several key aspects. Forecasts rely heavily on historical data and statistical methods to make educated guesses about future trends. Conversely, prediction involves making educated guesses or projections without depending on these factors.

However, both approaches share the most important assumption of predictive modeling: that past trends and patterns will continue into the future.

Although future results can be predicted and forecast based on past data, this is not always true. As such, businesses must continually evaluate and update their forecasting and forecasting processes to adapt to changing market dynamics and technological advancements.

That said, here are some of the main differences between prediction and forecast:

Definition

- Forecasting is the process of estimating future events or trends based on historical data and statistical methods. This involves analyzing patterns and trends in past data to make educated guesses about future outcomes.
- Prediction is making an educated guess or projection about a specific outcome without relying on historical data or statistical methods.

Temporal scale

- Temporal forecasting typically focuses on predicting outcomes over a longer period of time, often involving trends and patterns that occur over months, years, or even decades.
- Predictions can be more short-term and immediate, often used to estimate outcomes in the near future, up to a year.

Hypotheses

- Forecasting relies heavily on historical data and statistical methods, assuming that past trends and patterns will continue into the future.
- Predictions may involve assumptions based on expert opinions, hunches or subjective judgments, without necessarily relying on historical data.

Precision

- Forecasting aims to provide a more accurate estimate of future outcomes.
- Predictions may be less accurate due to their reliance on subjective judgments and assumptions.

Objective

- The forecast is used for planning strategy, budgeting, resource allocation and risk management. It helps businesses make informed decisions based on historical data and statistical projections.
- Predictions can be used in meteorology, sports predictions or market predictions, where subjective opinions or assumptions play a role.