1,721,265 research outputs found
Hou Tianya
No full text학위논문(석사)--아주대학교 일반대학원 :산업공학과,2009. 8I INTRODUCTION-------------------------------------- 1
II METHODS------------------------------------------- 5
II-1 Semi-Supervised Learning--------------------- 6
II-2 Technical Indicators Transform----------------- 8
II-3 Feature Extraction (PCA/NLPCA)--------------- 10
III EXPERIMENTS-------------------------------------- 15
III-1 Artificial Data---------------------------------- 16
III-2 Benchmark Data------------------------------- 20
III-3 Oil Price Data--------------------------------- 25
IV CONCLUSION--------------------------------------- 35
REFERENCES----------------------------------------- 36MasterOil price prediction is an important issue for the regulators of the government and the related industries. When employing the time series techniques for prediction, however, it becomes difficult and challenging since the behavior of the series of oil prices is dominated by quantitatively unexplained irregular external factors, e.g., supply- or demand-side shocks, political conflicts specific to events in the Middle East, and direct or indirect influences from other global economical indices, etc. Identifying and quantifying the relationship between oil price and those external factors may provide more relevant prediction than attempting to unclose the underlying structure of the series itself. Technically, this implies the prediction is to be based on the vectoral data on the degrees of the relationship rather than the series data. This paper proposes a novel method for time series prediction of using Semi-Supervised Learning that was originally designed only for the vector types of data. First, several time series of oil prices and other economical indices are transformed into the multiple dimensional vectors by the various types of technical indicators and the diverse combination of the indicator-specific hyper-parameters. Then, to avoid the curse of dimensionality and redundancy among the dimensions, the well-known feature extraction techniques, PCA and NLPCA, are employed. With the extracted features, a timepoint-specific similarity matrix of oil prices and other economical indices is built and finally, Semi-Supervised Learning generates one-timepoint-ahead prediction. The proposed method was validated on one artificial- and five real-world- problems. And then the series of crude oil prices of West Texas Intermediate (WTI) was used to verify the proposed method, and the experiments showed promising results: 0.86 of the average AUC and 88% of the average classification accuracy
A lab-on-chip sensor for in situ spectrophotometric measurement of seawater pH
No full textAccurate and high precision measurements of seawater pH are important for monitoring of ocean acidification and to fully understand the dynamics of the oceanic carbonate system. An optimisation and field testing of an in situ pH sensor, based on lab-on-chip technology, offers potential to address to this problem. The sensor employs a spectrophotometric method, with high precision (0.001 pH units) and accuracy (0.005 pH units) within the range of a certified TRIS buffer. In addition, it is a compact miniaturised field analyser that consumes minimal power (3 W) and uses low reagent consumption (2.2 μL per analysis). A newly developed Absorbance-Derivative method is used to eliminate indicator-induced pH perturbation; automatically correcting approximately ±0.08 pH units perturbation for each individual analysis. Another R-correction calibration method is also studied, resulting an accuracy of ±0.005 pH units, thereby providing an easy but also sufficiently accurate way to calibrate pH sensors. A recipe for preparing TRIS buffer with any pH value between 7.4 and 8.4, and with any salinity between 20 and 40 psu is resented, with initial verification suggesting an accuracy of ±0.004 pH units. The sensor has been field tested under various conditions, including vertical profiling, a brackish water moored deployment and a seawater moored deployment. A three-month pontoon deployment was performed, with an accuracy of ±0.003 pH units recorded over a 2 week period. However, biofouling in the inlet filter caused a diurnal offset up to 0.1 pH units, indicating that the next version of this pH sensor should focus on anti-biofouling technology
Le Sud et le monde nouveau
No full textJian Yu, Muyard Frank, Tianya. Le Sud et le monde nouveau. In: Perspectives chinoises, n°45, 1998. pp. 18-24
- …
