Cost-Efficient Predictive Auto-Scaling Using Transformer-LSTM Fusion Tuned with Bayesian Optimization
Cloud applications experience frequent and sometimes unpredictable shifts in demand due to user activity, daily usage cycles, and sudden workload spikes. Traditional autoscaling mechanisms used in cloud environments mostly follow reactive, threshold-based rules. They trigger scaling only after resources begin to satura...