Blockchain technology is emerging as one of the most enabling technologies of this century as it offers distinctive features like transparency, security and immutability. Consensus mechanism is a critical component to establish acceptance among the nodes on the current state of the blockchain. Various consensus protocols such as Proof of Work (PoW), Proof of Stake (PoS) are energy intensive and require extensive computational resources to bring agreement. As blockchain is revolutionizing various domains it also necessitates a novel consensus protocol to be build that will enhance efficiency through optimizations and integrations. This study proposes modified Proof of Reputation (mPoR), an optimized consensus mechanism based on dynamic scoring. mPoR can also effectively use various adaptive security methods to deter any malicious behavior in a dynamic peer to peer network. The pre-diction of network congestion and decentralization of consensus process can be achieved by integration of Reputation Dynamics Optimization (ReDO) which is a machine learning optimization method. mPoR along with ReDO is a promising innovation as applications requiring scalable, secure and high-performance decentralized systems can adopt this mechanism to enhance the overall efficiency. The incentive mechanism allows the nodes of the network with higher reputation to continue the engagement and boost reliability of the blockchain. The findings of the research is paramount as it can be utilities in new fields that require high efficiency, scalability and security.