Closed motifs for streaming time series classification

A streaming time series is a continuous and unbounded group of chronological observations that are found in many scientific and business applications. Motifs that are frequent subsequences are highly representative for the time series and play an important role in time series mining. Discovering motifs in time series has received much attention during recent years, and several algorithms have been proposed to solve this problem. However, these algorithms can only find motifs with a predefined length, which greatly affects their performance and practicality. Recent algorithms can discover motifs with different lengths, but require multiple scanning of the time series and are thus not applicable to streaming time series. In addition, it is difficult to determine the optimal length of interesting motifs; a suboptimal choice results in missing the key motifs or having too many redundant motifs. To overcome this challenge, we introduce the notion of a \(closed\) motif; a motif is \(closed\) if there is no motif with a longer length having the same number of occurrences. We propose a novel algorithm \(closedMotif\) to discover closed motifs in a single scan for streaming time series. We also use the nearest neighbor classifier with the most distinctive closed motifs to validate their potential in time series classification. Extensive experiments show that our approach can efficiently discover motifs with different lengths. In addition, our closed-motif-based classifier is shown to be more accurate than \(Logical\text{- }Shapelet\) , a state-of-the-art time series classifier. Finally, we demonstrate the scalability of \(closedMotif\) on several large datasets in diverse domains like video surveillance, sensor networks, and biometrics.     

On a Mixture GARCH Time‐Series Model

Abstract. Recently, there has been a lot of interest in modelling real data with a heavy‐tailed distribution. A popular candidate is the so‐called generalized autoregressive conditional heteroscedastic (GARCH) model. Unfortunately, the tails of GARCH models are not thick enough in some applications. In this paper, we propose a mixture generalized autoregressive conditional heteroscedastic (MGARCH) model. The stationarity conditions and the tail behaviour of the MGARCH model are studied. It is shown that MGARCH models have tails thicker than those of the associated GARCH models. Therefore, the MGARCH models are more capable of capturing the heavy‐tailed features in real data. Some real examples illustrate the results.     

The influence of sulfur dioxide on propane oxidation activity over supported platinum and palladium

Earlier studies have shown that sulfur dioxide and metal-support interaction can strongly influence propane oxidation over platinum. In particular, oxidation activity is enhanced when platinum is supported on sulfated -alumina or zirconia compared to -alumina. Therefore, it is of interest to compare the performance of palladium under the same experimental conditions. Four model catalysts were examined: Pt/-alumina, Pt/zirconia, Pd/-alumina and Pd/zirconia. The metal loading was kept at or below 0.05 wt% to emphasize changes in activity attributable to metal-support interaction. Reaction rates were measured with and without sulfur dioxide. Surface sulfation was analyzed by measuring acid strength and evaluating spectra obtained by Fourier-transform infrared spectroscopy. In contrast to platinum, sulfation does not promote propane oxidation on Pd/-alumina, and Pd/zirconia is less active than Pd/-alumina.     

Silane functionalization of perfluoroether oligomers for reaction management and morphology control of two‐phase epoxy networks

Mixtures of an epoxy resin, hardener, and acid functionalized perfluoroether oligomers will readily undergo phase separation during curing. However, the conditions to bring about the growth of nuclei into microscopic particles have hitherto been found only for systems cured with anhydrides. In the present study perfluoroether oligomers were functionalized by established procedures to introduce both carboxylic acid groups and alkoxysilane groups in sites within the chain extended segments. The presence of alkoxysilane groups together with the prereaction step with an excess epoxy resin, prior to the addition of the aromatic amine hardener, induced phase separation by a nucleation‐and‐growth mechanism. The dual functionality in the perfluoroether oligomer was even more beneficial when the alkoxysilane groups were hydrolyzed prior to the addition of the amine hardener. Under such circumstances the precipitation of the perfluoroether oligomer occurred quantitatively, as indicated by the complete absence of any plasticization effects in the epoxy matrix. From electron microscopy examinations, thermal analysis, and measurements of mechanical properties it was possible to deduce a plausible mechanism for the formation of the typical core‐shell aggregates within the precipitated particles for these systems, which could also be applied to other systems, such as those using carboxylic‐acid‐terminated butadiene acrylonitrile oligomers. © 2005 Wiley Periodicals, Inc. J Appl Polym Sci 98: 1036–1049, 2005     

A resilience optimization approach for workforce-inventory control dynamics under uncertainty

The presence of uncertainties in manufacturing systems and supply chains can cause undesirable behavior. Failure to account for these in the design phase can further impair the capability of systems to respond to changes effectively. In this work, we consider a dynamic workforce-inventory control problem wherein inventory planning, production releases, and workforce hiring decisions need to be made. The objective is to develop planning rules to achieve important requirements related to dynamic transient behavior when system parameters are imprecisely known. To this end, we propose a resilience optimization model for the problem and develop a novel local search procedure that combines the strengths of recent developments in robust optimization technology and small signal stability analysis of dynamic systems. A numerical case study of the problem demonstrates significant improvements of the proposed solution in controlling fluctuations and high variability found in the system’s inventory, work-in-process, and workforce levels. Overall, the proposed model is shown to be computationally efficient and effective in hedging against model uncertainties.     

Partition coefficient of diglycerides in crystallization of palm oil

Crystallization of palm oil with and without solvent was carried out over a temperature range of 10–25°C. The yields of olein, the diglyceride contents, and compositions of the stearin and olein phases were determined by thin-layer and gas-liquid chromatography. The three major diglycerides, analyzed as C₃₂, C₃₄, and C₃₆, are mainly dipalmitoyl glycerol, palmitoyloleoyl glycerol, and dioleoyl glycerol. In crystallization without solvent, C₃₂ (PP) had a strong affinity for the stearin fraction and C₃₆ diglycerides concentrated in the olein phase. The partition coefficient of diglycerides between the olein and stearin phases was temperature-dependent and was influenced by the type of solvent used. Although solvent enhances the diglyceride partition into the olein phase, partitioning is more effective at low temperatures and with acetone as the solvent for fractionation.     

Effect of water on HDS of DBT over a dispersed Mo catalyst using in situ generated hydrogen

A novel process was developed for the bitumen emulsion upgrading, wherein emulsion breaking and upgrading occurred in the same reactor using H₂ generated in situ from the water in the emulsion via the water gas shift reaction (WGSR). In this study, dibenzothiophene (DBT) was chosen as a model compound to investigate the effect of water and in situ H₂ on hydrodesulfurization (HDS). All the experiments were performed in a 1-L autoclave reactor at temperatures between 300 and 380 °C using in situ H₂ and ex situ H₂ (externally supplied H₂) over a dispersed Mo catalyst formed from phosphomolybdic acid (PMA). At very low water content, water was found to promote the HDS reaction in the ex situ H₂ run probably because it facilitates the formation of more active dispersed MoS_x species. At higher water content, however, water inhibits every individual reaction in the reaction network in the HDS of DBT, blocking the hydrogenation pathway more than the hydrogenolysis pathway. The relative reactivity of the in situ and ex situ H₂ depends on the water content present in the reaction system. At an optimized mole ratio of H₂O:CO (1.35), higher HDS activity was observed in the in situ H₂ run compared to ex situ H₂ run, and particularly, the hydrogenation pathway was promoted in the in situ H₂ run.     

Analysis of epoxidized natural rubber. A comparative study of d.s.c., n.m.r., elemental analysis and direct titration methods

A comparative analysis of epoxidized natural rubber samples by ¹H and ¹³C n.m.r., titrimetric, elemental and d.s.c. techniques has been made. Whereas the titrimetric method is only applicable at low epoxy contents (< 15 mol%) both n.m.r. methods give reasonable precision over the compositional range of 20–75 mol%. Elemental analysis appears less reliable. D.s.c. analysis through measurement of T_g provides the highest precision of measurement but requires independent calibration by one or more of the primary methods. The epoxy content may also be related to the polymer density.     

A study of the kinetics of nucleation in a palm oil melt

Nucleation from a supercooled melt of palm oil was studied by optical microscopy and differential scanning calorimetry (DSC). Despite being a multicomponent system, palm oil exhibits a rather simple cooling thermogram with its high- and low-T exotherms exclusively related to the “hard” and “soft” components of the oil. As the “hard” components are being removed, the position of the high-T peak shifts down toward the low-T peak with diminishing peak intensity, while the position of the latter remains virtually unchanged. At 288°K, nucleation in a palm oil melt is instantaneous. Its induction time-temperature curve shows an abrupt discontinuity at 297°K, which demarcates the occurrence of one polymorph from another. Nucleation data fit very well into the Fisher-Turnbull equation. Its larger activation free energy of nucleation is accompanied by lowering of the melting point and an increase in the crystal/melt interfacial free energy as compared to palm stearin. The slow rate of nucleation in palm oil is attributed to intermolecular interaction between its “hard” and “soft” components. Partly presented at the 1989 PORIM International Palm Oil Development Conference, September 5–9, Kuala Lumpur, Malaysia.