Robust M-estimators and Machine Learning Algorithms for Improving the Predictive Accuracy of Seaweed Contaminated Big Data

A common problem in regression analysis using ordinary least squares (OLS) is the effect of outliers or contaminated data on estimates parameters. robust method that not sensitive to and can handle needed. In this study, objective determine significant parameters moisture content seaweed after drying develop a hybrid model reduce outliers. The were collected with sensors from v-Groove Hybrid Solar Drier (v-GHSD) at Semporna, South-Eastern Coast Sabah, Malaysia. After second order interaction, we have 435 parameters, each parameter has 1914 observations. First, used four machine learning algorithms, such as random forest, support vector machine, bagging boosting by selecting 15, 25, 35 45 Second, developed methods M. Bi-Square, Hampel Huber. results show there improvement reduction number better prediction for big data. For highest variable importance seaweed, M Bi-square performs because it lowest percentage 4.08 %.