With the advancement of the Internet of Things (IoT), conventional principles of spectrum allocation cannot mitigate spectrum depletion so that a cognitive radio technology is proposed as a solution. The hidden primary user (PU) problem, however, is a critical issue in cognitive radio networks, because spectrum sensing nodes (SNs) can misclassify spectrum occupancy. To cope with this, machine learning-based cooperative spectrum sensing schemes (CSSs) have been proposed. The CSSs without considering the node placement, however, are still faced with the hidden PU problem. In this paper, we present how to place SNs to guarantee the performance of machine learning-based CSSs. We verify that the hidden PU problem makes the overlap of data distribution, which deteriorates the spectrum sensing ability. Based on the Kullback-Leibler divergence, analytical expressions for the spectrum sensing coverage of a single SN are derived. Then, we propose a strategy on how to place a few SNs to cover the whole area of the PU and prove the feasibility of our proposal by the experiment results.