The free trial version of Easy Predictive Analytics is valid for 7 days from the start of use.
If you do not sign up for a paid subscription after 7 days, you cannot use the Easy Predictive Analytics software. Please note that the free trial version will not automatically transition to a paid version.
Easy Predictive Analytics is compatible with Windows 10, 11 and Mac OS. Easy Predictive Analytics software can be accessed from Chrome and Edge browsers.
Easy Predictive Analytics requires tabular data. The data should include a column that will serve as the target for prediction (for example, a “Purchase” column indicating whether a purchase was made). Specify this column as the target when performing predictive data analytics. The predictive data analytics tool includes many sample datasets for reference.
Yes, it can. You call the predictive function from the API by using the API function.
In general, there are two ways to improve accuracy:
One is to increase the amount of data.
The more data you can provide for training, the higher the accuracy.
The second is to add columns to the data.
Adding columns that are likely related to the target for prediction can also improve the accuracy.
A dataset can have a maximum of 1 million rows. The limit is 10,000 rows for the time series mode. A dataset can have a maximum of 200 columns.
No, Easy Predictive Analytics automatically performs data preprocessing.
Preprocessing tasks like handling missing values or normalizing data are not required.
Easy Predictive Analytics uses neural networks and gradient-boosting trees. Each model has settings, with the optimal settings automatically selected based on the data.
For classification, Easy Predictive Analytics calculates a probability score as the prediction value, whereas for regression, it calculates a numerical value. If one item is excluded, the numerical value is reduced. The larger this difference, the more it affects the prediction. This is considered a contribution. The contribution of each item is calculated by comparing the prediction without using the item with the prediction using all the items. For more information on the calculation of contribution, click here.
The training duration varies depending on the data.
It takes about 4 hours to train for the recommended data size limit of 1 million rows x 200 columns.
If the number of data points is 500 or less, cross-validation is performed irrespective of whether the “Always perform cross-validation” checkbox is checked. Cross-validation will not be performed if you specify the evaluation data, as only the specified data will be used for evaluation.
A cancellation link is provided in the email sent by Paddle when you purchase a license. Please use that link to cancel your subscription.
A link to change the payment method is provided in the email sent by Paddle when you purchase a license. Please use that link to change your payment method.
Easy Predictive Analytics has been provided as part of a limited sales period since December 2024. The limited sales period may end in approximately one year. If we decide to terminate the service, we will notify you at least three months in advance. Support will be provided during this period as well. If you have any questions or issues, please contact our support team.
