Cloud Resource Bundle Selection Optimizer

The patent segments the resource selection process into multiple filtering stages. First, combinations are filtered by hard constraints (budget, minimum resource requirements). Then, the remaining combinations are evaluated using a scoring function that considers multiple factors (compute power, cost, distribution). This segmentation allows the system to handle large search spaces efficiently by eliminating obviously suboptimal combinations early in the process.

The patent applies partial action by not evaluating all possible combinations exhaustively. Instead, it uses filtering criteria to identify a subset of promising combinations and evaluates only those. The scoring function provides a heuristic that guides the selection process, allowing the system to achieve good enough solutions without complete enumeration, thus reducing computational overhead while maintaining high selection accuracy.

The patent changes the parameters of resource allocation by introducing a scoring function that weights different factors (compute power, cost, distribution) according to user preferences and budget constraints. This allows the system to dynamically adjust allocation decisions based on the relative importance of different parameters, finding the optimal balance between satisfying workload demands and staying within budget limits.

The patent implements dynamic resource selection where the filtering and scoring criteria can be adjusted based on changing conditions. The system can adapt to different budget constraints and workload requirements by modifying the filtering thresholds and scoring weights, enabling flexible allocation that responds to real-time conditions while maintaining optimality.

The patent segments the filtering process into distinct stages: first applying hard constraints (budget limits, minimum resource requirements) to eliminate obviously invalid combinations, then applying the scoring function to evaluate remaining combinations. This segmentation allows the system to apply complex evaluation criteria only to a reduced set of candidates, maintaining high selection accuracy while minimizing processing time.

The patent performs preliminary filtering using simple hard constraints before applying the more complex scoring function. By pre-filtering combinations that clearly violate budget or minimum requirement constraints, the system reduces the search space for subsequent evaluation, thereby reducing overall processing time while maintaining accurate selection through the comprehensive scoring of remaining candidates.