An Efficient Medicine Demand Prediction System Using LTH-SES-Based Machine Learning Technique with Pharmacy Supply Chain

Abstract

The Pharmaceutical Industry (PI) is deemed as one amongst the most substantial industrial sectors. Therefore, for the healthcare system, effective management of the Pharmaceutical Supply Chain (PSC) is crucial. Thus, an efficient medicine demand prediction system is proposed in this paper by using the Logistic Tanh activation adapted Single Exponential Smoothing (LTH-SES)-based Machine Learning (ML) technique with PSC. The previous medication sales data from the PSC are collected from publicly available sources to predict future medication needs. Later, to reduce the vast amount of information present in the dataset, the MapReduce model is performed. Later, the features are extracted. After that, by using the Renyi Entropy Principal Component Analysis (REPCA) technique, the dimensionality is reduced. Later, by using Min-Max Distance Centroid Fuzzy C-Means (M²DCFCM) clustering, the medications are grouped together. Lastly, for forecasting the future demand for medications, the LTH-SES technique is used. The proposed system’s performance is further validated and the outcomes exhibited that the proposed methodology forecasts the demand more effectively than other prevailing techniques.