Since the adoption of the Blockchain network by modern technologies, which benefit from its immutability and secure access provided by its consensus process, blockchain mining has gained popularity. Blockchain networks’ miners rely on the consensus procedure to contribute data to the network. Miners use their resources to publish new blocks and receive rewards, and they participate in the mining process. Since solo mining might be difficult, most miners decided to join a mining pool. Several issues surrounding the miner’s choice with pool mining remain unanswered because of the range of many mining pools and the potential adoption of various incentive systems. There are a few issues to consider, though, about maintaining a miner’s trust throughout the consensus and mining processes. As part of the consensus, miners that are part of a mining pool contribute their processing power toward the task of adding a block. The pool will be rewarded if they are successful in their endeavors. To guarantee the execution of the consensus protocol in pool mining, a trust evaluation model is presented in our study. As a result, trust assessment among miners is formulated in this research it is evaluated as a classification problem, and an innovative approach using machine learning is offered to classify trust. First, the vector for the trust features is built using the parameters of the Blockchain. A trust classifier can then be created by training it with samples of data that miners have acquired and that contain transaction activities. Also, the provided Trust model verifies that the total computing power of the pool for which miners are requested and the miner’s computation power should be 50 percent greater than or equal to the total computation power this is for to mitigate attacks brought on by untrusted miners through forks in the Blockchain. The Hyperledger tool was used as the framework to assure robustness in analyzing transaction delay, and transaction throughput, and to achieve improved Blockchain performance. The research was carried out to demonstrate the viability of our Trust Model and to show that the trained trust classifier has comparatively high processing power.