The Internet of Medical Things (IoMT)-based Remote Patient Monitoring (RPM) systems provide real-time data and insights about patients' conditions without the need for constant physical visits to healthcare facilities. The wearable sensors are responsible for sensing the psychological vitals of humans at different intervals to detect their precise health status. This article introduces an Interfering Input Classification Model (IICM) for addressing the continuous and discrete data extraction issues in RPM. The proposed model hybridizes whale optimization and deep-Q-learning for classification and connectivity identification between different interval data. The sensed information is segregated as discrete and continuous using the search process termination of the whales. Such classified data is validated for the maximum discreteness and its interconnection between the previous sequences. This interconnection is performed using Q-learning based on different state changes through continuous clinical range correlations. The proposed model exemplifies continuous and discrete signals (data) for abnormality detection and diagnosis recommendation. Thus the proposed model is reliable in improving the classification ratio and accuracy with fewer errors. The experimental values of the proposed model are as follows: IICM enhances classification by 6.65%, signal identification by 6.63%, and classification accuracy by 10.08%. This model reduces classification time by 9.51% and error by 10.14% for different ranges of inputs.