5G Handover Prediction System

Engineered a machine learning pipeline using TensorFlow, Scikit-learn, Pandas, and NumPy to process 103K+ 5G network samples.Deployed an LSTM signal predictor and a CNN-BiLSTM hybrid classifier to predict cellular handover trigger events.Resolved a 47:1 class imbalance using Temporal SMOTE with 50-timestep sliding windows, achieving 98.37% prediction accuracy.