An Adult Care Facility (ACF) is a healthcare organization providing regular non-medical services to the disabled elderly people. The number of ACF and elderly care homes are rising in US. Forecasting the number of people in a facility based on other factors can be very useful for planning and scheduling resources. In this paper, machine learning models are developed with the purpose of predicting the total number of patients admitted in an ACF at the end of each quarter. In particular, this paper proposes two models: Linear Regression (LR) and Deep Neural Network (DNN). Both models were used to fit the quarterly data obtained from multiple ACFs. The performance of the models was evaluated by using the known R-squared score. Based on R-squared scores on training, validation, and testing, the LR model outperformed the complex DNN model as the best prediction model.
Online shoppers usually look for their desired products with maximum discount possible, as larger discounts tend to influence the consumer behavior. However, the price of a product also plays significant role in buying decision because of the limited budget. In this paper, to optimize the shopping decision, a 0/1 knapsack problem is formulated, and a comparison of the traditional recursive solution method and the dynamic programming optimization approach is made. The results show that the dynamic programming method outperforms the recursive solution method in terms of computational time for large number of purchase items.