Unmanned aerial vehicles (UAVs) are expected to be extensively used as an integral part in the future generations of communication networks, to provide ubiquitous connectivity. The mobile nature of UAVs make them a tempting candidate to provide seamless connectivity in environments where the installation of conventional terrestrial base stations (BS) is not feasible. Nonetheless, there are major deployment issues related to optimal placement of UAV-mounted base stations (UAV-BSs) due to limited number of UAV-BSs, limited energy availability and trade-off between coverage area and its altitude. In this paper, we address UAV-BSs placement issues by proposing a novel Machine learning (ML) based intelligent deployment mechanism. More specifically, for intelligent deployment of UAV-BSs based on energy, computational power, nature of available data and criticality of the scenario, we use two different approaches: Support Vector Machine (SVM) and Deep Learning (DL), which is composed of sequential time series learning process. Moreover, to address the security and privacy challenges emanating from the wireless connectivity and untrusted broadcast nature of UAV-BSs, we propose a Blockchain-based novel information-sharing scheme. To evaluate the performance of our combined secure and intelligent proposed approach, we have improved energy consumption by almost twice in contrast with the normal deployment of UAV-BSs.