In cloud computing settings, where processes or applications operating on virtual machines or containers demand computational resources like CPU, memory, and storage, efficient resource allocation is essential for maximising performance. In cloud computing, performance optimisation means getting high throughput, short response times, and best use of available resources to guarantee end users have an acceptable application experience while keeping cloud service providers' costs down. This research explores sophisticated load scheduling strategies to improve cloud computing efficiency. These approaches go beyond conventional ways to make intelligent judgements regarding resource allocation, and they include dynamic, heuristic-based, machine learning-based, and adaptive scheduling approaches. Dynamic scheduling methods make real-time modifications to ensure optimum resource allocation based on current demand by continually monitoring system conditions and workload characteristics. Heuristic-based scheduling finds near-optimal solutions in difficult situations by using heuristic techniques like genetic algorithms and ant colony optimisation. With the use of past data and machine learning methods, machine learning-based scheduling is able to forecast future workload patterns and dynamically optimise resource allocation, responding to changing workload situations over time. Adaptive scheduling strategies provide adjustments to resource allocation in response to system input and shifting workload circumstances. This guarantees that resources are distributed effectively to fulfil performance standards and accommodate demand variations. The efficiency of cloud computing systems is greatly increased by enhanced load scheduling algorithms, which also improve system performance and resource utilisation. These methods have advantages, but they also have drawbacks, such large computational overhead, complicated implementation, and ongoing workload pattern adaption. To further enhance resource allocation and performance in cloud computing settings, future research initiatives may concentrate on tackling these issues, creating more effective algorithms, and investigating cutting-edge techniques.