Share ideas related to machine learning use cases in telecom

[longtermevolution]

In this thread you can all share your ideas for different use cases of machine learning using python in telecom.

probably good to get ideas which can be convert later to proper model.


Cheers!!!

[longtermevolution]

1. Anomaly detection based on performance 4g KPIs.
2. Identify and score all cells in network based on capacity, Quality, revenue(3 layers). This way we can choose sites for capacity expansion.
3. Coverage Hole Prediction(Based on that can plan new sites)
4. Dynamic Load balancing in network

[amer1]

It's a good idea for the AI to propose what radio parameter to change based on the KPI performance. Like for example in case of dropped calls, the AI can propose the different points to check first and then propose what parameters needed to change to solve the problem.

[oswe]

using simple machine learning you can do linear regression using as inputs PRB_utilization , RRC users, and estimate or predict user_throughput. Ive been testing and is ok