Search

Nowadays, in the topics related to prediction, in addition to increasing the accuracy of existing algorithms, the reduction of computational time is a challenging issue that has attracted much attention. Since the existing methods may not have enough efficiency and accuracy, we use a combination of machine-learning algorithms and statistical methods to solve this problem. Furthermore, we reduce the computational time in the testing model by automatically reducing the number of trees using penalized methods and ensembling the remaining trees. We call this efficient combinatorial method “ensemble of clustered and penalized random forest (ECAPRAF)”. This method consists of four fundamental parts. In the first part, k-means clustering is used to identify homogeneous subsets of data and assign them to similar groups. In the second part, a tree-based algorithm is used within each cluster as a predictor model; in this work, random forest is selected. In the next part, penalized methods are used to reduce the number of random-forest trees and remove high-variance trees from the proposed model. This increases model accuracy and decreases the computational time in the test phase. In the last part, the remaining trees within each cluster are combined. The results of the simulation and two real datasets based on the WRMSE criterion show that our proposed method has better performance than the traditional random forest by reducing approximately 12.75%, 11.82%, 12.93%, and 11.68% and selecting 99, 106, 113, and 118 trees for the ECAPRAF–EN algorithm.

As the risk of diseases diabetes and hypertension increases, machine learning algorithms are being utilized to improve early stage diagnosis. This study proposes a Hybrid Prediction Model (HPM), which can provide early prediction of type 2 diabetes (T2D) and hypertension based on input risk-factors from individuals. The proposed HPM consists of Density-based Spatial Clustering of Applications with Noise (DBSCAN)-based outlier detection to remove the outlier data, Synthetic Minority Over-Sampling Technique (SMOTE) to balance the distribution of class, and Random Forest (RF) to classify the diseases. Three benchmark datasets were utilized to predict the risk of diabetes and hypertension at the initial stage. The result showed that by integrating DBSCAN-based outlier detection, SMOTE, and RF, diabetes and hypertension could be successfully predicted. The proposed HPM provided the best performance result as compared to other models for predicting diabetes as well as hypertension. Furthermore, our study has demonstrated that the proposed HPM can be applied in real cases in the IoT-based Health-care Monitoring System, so that the input risk-factors from end-user android application can be stored and analyzed in a secure remote server. The prediction result from the proposed HPM can be accessed by users through an Android application; thus, it is expected to provide an effective way to find the risk of diabetes and hypertension at the initial stage.

A carsharing service can be seen as a transport alternative between private and public transport that enables a group of people to share vehicles based at certain stations. The advanced carsharing service, one-way carsharing, enables customers to return the car to another station. However, one-way implementation generates an imbalanced distribution of cars in each station. Thus, this paper proposes forecasting relocation to solve car distribution imbalances for one-way carsharing services. A discrete event simulation model was developed to help evaluate the proposed model performance. A real case dataset was used to find the best simulation result. The results provide a clear insight into the impact of forecasting relocation on high system utilization and the reservation acceptance ratio compared to traditional relocation methods.