I'm working on refining my forecasting models in IBM Planning Analytics and wanted to reach out to see how others approach long-term predictions. I'm particularly interested in best practices for managing complex data sets and ensuring the models stay accurate over time.

For instance, when working with volatile data like cryptocurrency trends, building reliable models can be quite challenging. I've been exploring methods to improve my forecasts, especially for something as unpredictable as the Bitcoin price prediction 2025. Are there any specific tools, techniques, or even built-in features within IBM Planning Analytics that you've found helpful for similar long-term forecasts?

I'd love to hear how you manage uncertainty and volatility in your models, as well as any tips for optimizing performance when dealing with large datasets.

------------------------------
Ivvyy Risea
------------------------------

Hello Ivvyy,

As you can find in the documentation here, Planning Analytics has several built-in forecasting functions that may be suitable for your case. However, if you have access, my best recommendation is to use the IBM Granite Timeseries Foundation Model from watsonx.ai, as described here.

It will require some Python programming skills,  but you can integrate input and output through the TM1 REST API (documentation) to leverage the Granite Timeseries Foundation Model from watsonx.ai.

Let me know if you need any help!

------------------------------
Thiago Teixeira
IBM Champion
CTO - RCI Analytics Intelligence
------------------------------

Original Message:
Sent: Sun February 16, 2025 05:30 AM
From: Ivvyy Risea
Subject: Best Practices for Forecasting Models in IBM Planning Analytics

I'm working on refining my forecasting models in IBM Planning Analytics and wanted to reach out to see how others approach long-term predictions. I'm particularly interested in best practices for managing complex data sets and ensuring the models stay accurate over time.

For instance, when working with volatile data like cryptocurrency trends, building reliable models can be quite challenging. I've been exploring methods to improve my forecasts, especially for something as unpredictable as the Bitcoin price prediction 2025. Are there any specific tools, techniques, or even built-in features within IBM Planning Analytics that you've found helpful for similar long-term forecasts?

I'd love to hear how you manage uncertainty and volatility in your models, as well as any tips for optimizing performance when dealing with large datasets.

------------------------------
Ivvyy Risea
------------------------------

Hello Ivvy, 

A couple of answers / references to help here: 

• As Thiago references, Python with Planning Analytics can assist, using TM1py and more details here for timeseries forecasting https://code.cubewise.com/blog/timeseries-forecasting-with-facebook-prophet-and-tm1-planning-analytics/#cpsnh2-0
• Related to Watsonx and Granite, similar and related to Thiago's comments. Tomorrow will be presenting at Cubewise Innovation Week 2025 and demonstrating some of these move advanced topics to help further start the conversation https://events.teams.microsoft.com/event/45a81e4e-3b2b-4e74-b43b-0b98e4dd7753@0635d657-0279-4110-9c8f-2ee27b1e065b

Regards

------------------------------
James Downey
General Manager
Cubewise
TX
------------------------------

Original Message:
Sent: Sun February 16, 2025 05:30 AM
From: Ivvyy Risea
Subject: Best Practices for Forecasting Models in IBM Planning Analytics

I'm working on refining my forecasting models in IBM Planning Analytics and wanted to reach out to see how others approach long-term predictions. I'm particularly interested in best practices for managing complex data sets and ensuring the models stay accurate over time.

For instance, when working with volatile data like cryptocurrency trends, building reliable models can be quite challenging. I've been exploring methods to improve my forecasts, especially for something as unpredictable as the Bitcoin price prediction 2025. Are there any specific tools, techniques, or even built-in features within IBM Planning Analytics that you've found helpful for similar long-term forecasts?

I'd love to hear how you manage uncertainty and volatility in your models, as well as any tips for optimizing performance when dealing with large datasets.

------------------------------
Ivvyy Risea
------------------------------