Recent research in Task Scheduling Optimization in Cloud Computing focuses on improving efficiency, scalability, and energy consumption through advanced heuristic, meta-heuristic, and learning-based algorithms. Studies integrate hybrid optimization techniques such as Particle Swarm Optimization, Grey Wolf Optimizer, and Cuckoo Search with reinforcement learning and deep neural models to achieve faster convergence and balanced resource utilization. These methods aim to minimize makespan, cost, and task waiting time while enhancing throughput and Quality of Service (QoS) under dynamic workloads. The latest approaches emphasize adaptive and intelligent scheduling frameworks that self-optimize resource allocation, ensuring high performance and energy efficiency in heterogeneous cloud environments.