Scheduling of a large number of submitted tasks is a central operation in cloud computing. Efficient scheduling and resource allocation for the submitted tasks ensures that Service-Level-Agreements (SLA) violations are mini-mized. We present a fuzzy logic-based approach for predicting submitted tasks which are likely to encounter SLA violations. It may help Cloud Service Providers (CSPs) to design corrective interventions in terms of additional resource allocation to prevent SLA violations. The proposed mechanism assists in reducing SLA violations and improves the end-user quality-of-service experience along with enhancement of CSP revenues. The appropriate selection of performance metrics has enabled the proposed model to achieve the highest classification accuracy of 92.6 percent in predicting SLA violation.