Promoted content


Smaller values suggest lower impact of early conditions on new predictions

The prediction and the underlying analysis requires you to know the number of variables and number of observations for each variable. If you are predicting the 9th number in variable A {5 1 5 1 5 2 5 1} based on your knowledge of variable B {3 4 3 4 3 2 4 2}, then you have 2 variables and 8 observations.

Unless you specify variables in the file are named as V1, V2, V3, V4, V5 ...
If you would like to name your variables, please click below.

What are the variable names in the data?

Please use single words to name variables. Example: 'Regional-sales' instead of 'Regional sales'.
If you have one variable, you should use the same variable name you have entered above.

Please upload your data file in a comma-separated file format

First row is for the earliest observation, last row is the latest.
First column is for the variable which is predicted by default, last column is the last variable that can explain the predicted variable. All data should be numeric, hence please exclude the time column and columns with characters.
File path:

Decision/ guessed variables and values for the predicted period

If your variables have names (computer-use only), it can be:
Number_of_machines 2; Load_in_machines 25
Please note that variable names are alphanumeric, please use single words for variable names

By using this service you agree and accept that is not responsible for the output values or interpretations of the output, any use purposes of the user, leakage of sensitive data beyond our best intentions to protect privacy. does not collect or store the entered data. We store variable names and survey responses in aggregate and only for scientific research purposes and improving the service quality. respects privacy of user data and has no interest in cookies. The provided service aims to enhance the analytic capabilities of users, but it is free of any guarantees. All responsibility belongs to the user.

Prediction model details:

Model fit

Model parameters

Observations vs. model results


Prediction boundaries are limit levels within the 95% prediction interval
We use 'TS' in short to say 'Time series'. These models require each observation to have the same frequency, e.g., weekly, monthly, quarterly, yearly observations
Conditional means do not require a temporal sequence
For variables that have constant growth rates, time trend is a necessary variable to account for that type of growth characteristic (#1, #2, #3, #4 and #5)
Seasonality effects is a common term used to account for periodic patterns that influence a variable, e.g., temperature and months (#2 #3 ,#4 and #5)
Dynamic effects arise when the current value of a variable depends on previously observed values of the same variable. (#3 and #5 in single variable and multivariable models)
Long term memory effects are observed when starting conditions have intertemporal effects (#4 and #5).

Thank you for predicting with us!