Due to the imbalanced nature of datasets, classifying unbalanced data classes and drawing accurate predictions is still a challenging task. Sampling procedures, along with machine learning and deep learning algorithms, are a boon for solving this kind of challenging task. This study’s objective is to use sampling-based machine learning and deep learning approaches to automate the recognition of rotting trees from a forest dataset. Method/Approach: The proposed approach successfully predicted the dead tree in the forest. Seven of the twenty-one features are computed using the wrapper approach. This research work presents a novel method for determining the state of decay of the tree. The process of classifying the tree’s state of decay is connected to the issue of unequal class distribution. When classes to be predicted are uneven, this frequently hides poor performance in minority classes. Using stratified sampling procedures, the required samples for precise categorization are prepared. Stratified sampling approaches are employed to generate the necessary samples for accurate prediction, and the precise samples with computed features are input into a deep learning neural network. Finding: The multi-layer feed-forward classifier produces the greatest results in terms of classification accuracy (91%). Novelty/Improvement: Correct samples are necessary for correct classification in machine learning approaches. In the present study, stratified samples were considered while deciding which samples to use as deep neural network input. It suggests that the proposed algorithm could accurately determine whether the tree has decayed or not.

As a cause of interfering with routine activities, freezing of gait (FOG) is a severe syndrome of Parkinson’s disease (PD) and usually performs as an abrupt and momentary inability to effective stepping forward. Advanced wearable acceleration sensors based on socially implemented Internet of medical things (IoMT) devices can remotely provide a platform for recognizing FOG. However, due to the diverse data acquisition modes that appear in classic IoMT devices, the obtained data may contain imprecise, hesitant, and incomplete ones. Meanwhile, the bounded rationality owned by neurologists usually has a big impact on using wearable acceleration sensors to predict illnesses. Therefore, the objective of this article lies in exploring a fuzzy intelligence learning approach based on bounded rationality in IoMT systems and providing a valid scheme for biomedical data analysis. Specifically, a brand-new three-way group decision-making approach by means of TODIM (an acronym in Portuguese for interactive multicriteria decision-making) with incomplete dual hesitant fuzzy (DHF) information and its applications in detecting FOG in PD using IoMT devices are systematically explored. First, taking advantage of DHF sets (DHFSs) when depicting realistic group decision information, the concept of multigranulation (MG) incomplete DHF information systems is built. Second, adjustable MG DHF probabilistic rough sets (PRSs) are further put forward via DHF similarity relations. Third, a three-way group decision-making approach is constructed by virtue of adjustable MG DHF PRSs and TODIM. Finally, the validity, effectiveness, and practicality of the constructed three-way group decision-making approach are investigated by a University of California, Irvine (UCI) dataset with several experimental analyses in the background of FOG detection in PD using IoMT devices. The experimental result indicates that the developed fuzzy intelligence learning approach achieves reasonable diagnostic conclusions for FOG detection in PD from the perspective of uncertain information processing abilities, decision risks, and bounded rationality.