Browsing by Author "Wang, Yau-Shian"

Now showing 1 - 2 of 2

Fouling mitigation of a dead-end microfiltration by mixing-enhanced preoxidation for Fe and Mn removal from groundwater
(2013-02-20)
Lin, Jr-Lin
;
Huang, Chihpin
;
Ruhsing, Pan Jill
;
Wang, Yau-Shian
;
環境工程研究所
A combination of oxidation and microfiltration (oxidation-MF) is commonly used to remove Fe and Mn from groundwater. The characteristics of Fe and Mn oxides formed by oxidation can be significantly affected by the mixing intensity during oxidation, which subsequently influences the performance of membrane filtration. To optimize oxidation-MF, the critical role of mixing intensity on Fe and Mn oxide formation was investigated. Sufficient oxidant (NaOCl) was added to a prepared ferrous solution with or without Mn ions to oxidize the ions at different mixing intensities, followed by a dead-end MF for 30 min at a constant transmembrane pressure of 1 kg/cm(2). A custom-made module surrounded by a tube membrane was used in microfiltration. The resistance of membrane filtration was calculated by a resistance-in-series model. Cake resistance (R-c) dominated the fouling of the membrane, which depended on the composition of oxides. Oxides formed at high mixing intensity were smaller and more compact, which resulted in reduced R-c, and increased permeate flux. The presence of oxidized Mn aggravated membrane fouling. Permeate flux declined rapidly as a result of the cake formed by Fe-Mn oxides with increased hydrophobicity. The irreversible fouling of the dead-end MF was insensitive to mixing intensity during prechlorination, while the irreversible fouling induced by Fe-Mn oxides was more severe than that by Fe oxides as membrane filtration cycle progressed. The results show that membrane flux of a dead-MF can be effectively improved by prechlorination with enhanced mixing due to the formation of a permeable cake layer with low cake compressibility in the membrane filtration. (c) 2012 Elsevier B.V. All rights reserved.
Multi-Source Learning for Sales Prediction
(2017-01-01)
Tsai, Kuen-Han
;
Wang, Yau-Shian
;
Kuo, Hsuan-Yu
;
Tsai, Jui-Yi
;
Chang, Ching-Chih
With the convenience and popularity of Internet, the sales on e-commerce platforms have grown significantly. This exponential growth generates massive chunks of data, which provides the opportunity to utilize the historical data for predicting the customers' behaviors and can thus offer better services. One of the major tasks is to correctly estimate the coming sales of products since the e-retailers can reserve the products in a smart way. However, even with massive data, it is still challenging to estimate the sale amount of each product due to 1) the complicated language structures, 2) difficulties in integrating the features, and 3) missing values. To address these issues, we propose a framework for predicting the sales, which contains two phases: 1) feature extraction from product attributes and reviews, and 2) tensor decomposition for multi-source learning. The experimental results show that our framework outperforms baselines by 73%.