About Advanced Predictions

Use the power of Advanced Predictions in FreeForm to improve your forecasts using multiple input drivers and sophisticated forecasting and machine learning algorithms.

Advanced Predictions are an extension of Auto Predict, expanding the power of the prediction engine by using machine learning backed by OCI Data Science Cloud. While Auto Predict and Predictive Planning offer univariate forecasting based on a single measure, Advanced Predictions provide multivariate predictions, looking at multiple input drivers to make the prediction. You train the ML model, which improves the quality of the predictions.

Univariate predictions used in Auto Predict and Predictive Planning create predictions based on one variable, for example, analyzing historical data for sales volume to predict future sales volume.

Multivariate predictions used in Advanced Predictions use advanced statistical and machine learning algorithms to generate multivariate predictions to improve your forecasting accuracy. Predictions are based on multiple variables. For example, to predict future sales volume, you can analyze several drivers or variables that correlate with your target variable. For example, you can analyze Industry Volumes, Average Sales Price, Advertising and Promotions, GDP Rate, and so on. The algorithms take the historical data for that variable and the drivers you included, as well as their what-if projections, to create a more accurate prediction.

With Advanced Predictions, you can predict using any of these algorithms or use Oracle AutoMLx, which runs all of these algorithms and selects the best option with the best results for you:

Light GBM
XGBoost
Prophet
SARIMAX

Administrators use the familiar IPM Configuration wizard to easily configure Advanced Predictions to provide insights into solutions.

Depending on the implementation, planners can use a dashboard to see an overview of the prediction, which can use multiple external and input drivers, such as industry volumes, advertising and promotions, sales price, and GDP rates.

Depending on the implementation, planners can also review:

Prediction—Review prediction details, historical data, and input driver details.
Input Drivers—Review and edit the input driver details.
Forecast Accuracy—Review and ensure the accuracy of the prediction.
Predict by Algorithms—Explore prediction results using different algorithms.

Explain Prediction opens prediction details that contain a line chart for historical trends and prediction results considering the best case, worse case, and most likely scenarios. You can also review details about the prediction including accuracy, the error measure, and the algorithm that was used.

Your current investment in Cloud EPM is delivering AI to you:

The AI features are already built into the business processes and is powered by OCI AI and leading large language models.
Your results are based on the data you provided and the relevant context surrounding it, making sure that your predictions are accurate and aligned with your specific data sources.
It honors Cloud EPM’s multi-layer security structure.
Its extensible framework allows for new capabilities and functionalities.

Advanced Predictions:

Enables more powerful predictions by leveraging more sophisticated algorithms
Improves accuracy as it correlates with multiple provided data points
Easy to configure and use with the IPM Configuration wizard that can automatically select the best algorithm to use
Designed for business finance, embedded in Cloud EPM so it can be used in context, without needing to learn data science

Videos

Your Goal	Watch This
Learn how Advanced Predictions can perform AI-assisted data analysis that helps you streamline reporting and strengthen decision making. Advanced Predictions takes advantage of machine learning models to generate multivariate predictions to improve your forecasting accuracy.	Introducing Advanced Predictions in Cloud EPM