Machine learning and AI in Telecom

ML Project end to end – RAN

Machine learning (ML) is a field of study that gives computers the ability to learn without being explicitly programmed.

Here, end to end Machine learning project implementation is captured for RAN predictive analytics in Telecom Industry using Jupyter notebook, Python programming, Spyder IDE and anaconda data science/ML platform.

Key steps during ML project implementation -

a. Data collection – Raw data is collected from NMS here basis the defined objective of doing predictive analytics of “throughput” based on inputs – PRB usage, RRC conn. Users, and MCS

b. Data preparation – Refined the raw report by extracting only required features (RAN parameters/counters) from the report, remove null, missing values, convert the units wherever required (e.g. – kbps to Mbps in throughput)

c. Data visualization and ML Model selection – Used Jupyter notebook to visualize the data to understand the features and its correlation/importance. Then choose the ML model depending upon the relation between input and output features. Here, random forest regressor is used.

d. Cross validation and training of ML Model - Dataset is divided into train and test data to test the model's generalization ability using out-of-the-sample data chunks. Here “Randomized Search CV” is used as data has low dimensions.

e. Evaluate the ML Model – Check the predicted and actual values by measuring cost function (mean square error) here. Difference between predicted and actual value shows a normal distribution plot, which signifies model is okay.

f. Predictive Analytics – IDE is used to create app using html files, whereby adding any new input values (for PRB usage, RRC conn. Users, MCS), predicted output (throughput) is generated.

Check more details on Machine learning through complete course -https://www.itelcotech.com/learningpath/ml-ai-in-telecom-networks