Improving bagging performance through multi-algorithm ensembles

Working as an ensemble method that establishes a committee of classifiers first and then aggregates their outcomes through majority voting, bagging has attracted considerable research interest and been applied in various application domains. It has demonstrated several advantages, but in its present form, bagging has been found to be less accurate than some other ensemble methods. To unlock its power and expand its user base, we propose an approach that improves bagging through the use of multi-algorithm ensembles. In a multi-algorithm ensemble, multiple classification algorithms are employed. Starting from a study of the nature of diversity, we show that compared to using different training sets alone, using heterogeneous algorithms together with different training sets increases diversity in ensembles, and hence we provide a fundamental explanation for research utilizing heterogeneous algorithms. In addition, we partially address the problem of the relationship between diversity and accuracy by providing a non-linear function that describes the relationship between diversity and correlation. Furthermore, after realizing that the bootstrap procedure is the exclusive source of diversity in bagging, we use heterogeneity as another source of diversity and propose an approach utilizing heterogeneous algorithms in bagging. For evaluation, we consider several benchmark data sets from various application domains. The results indicate that, in terms of F1-measure, our approach outperforms most of the other state-of-the-art ensemble methods considered in experiments and, in terms of mean margin, our approach is superior to all the others considered in experiments.     

Effect of Monovalent Cation Additives on the γ-Al2O3-to-α-Al2O3 Phase Transition

The effect of monovalent cation addition on the γ-Al₂O₃-to-α-Al₂O₃ phase transition was investigated by differential thermal analysis, powder X-ray diffractometry, and specific-surface-area measurements. The cations Li⁺, Na⁺, Ag⁺, K⁺, Rb⁺, and Cs⁺ were added by an impregnation method, using the appropriate nitrate solution. β-Al₂O₃ was the crystalline aluminate phase that formed by reaction between these additives and Al₂O₃ in the vicinity of the γ-to-α-Al₂O₃ transition temperature, with the exception of Li⁺. The transition temperature increased as the ionic radii of the additive increased. The change in specific surface area of these samples after heat treatment showed a trend similar to that of the phase-transition temperature. Thus, Cs⁺ was concluded to be the most effective of the present monovalent additives for enhancing the thermal stability of γ-Al₂O₃. Because the order of the phase-transition temperature coincided with that of the formation temperature of β-Al₂O₃ in these samples, suppression of ionic diffusion in γ-Al₂O₃ by the amorphous phase containing the added cations must have played an important role in retarding the transition to α-Al₂O₃. Larger cations suppressed the diffusion reaction more effectively.     

Gas chromatographic identification of fatty acids,fatty alcohols,and hydrocarbons ofHibiscus rosa-sinensis leaves

Hibiscus rosa-sinensis leaves (family Malvaceae) were analyzed for their fatty acid, fatty alcohol, and hydrocarbon contents. Wax hydrocarbons ranging from C₁₆ to C₃₂ with C₂₃, C₂₅, C₂₇, and C₃₁ as major components and wax alcohols between C₂₁ and C₃₀ with C₂₆, iso-C₂₈, and iso-C₃₀ as major components were found to be present in the petroleum ether fraction of the leaves. Fatty acids ranging from C₈ to C₂₈ with C₈, C₁₂, C₁₄, C₁₆, and C_18:2 as major components were found in the combined form. Two cyclic acids, sterculic and malvalic, have also been identified.     

Determination of interfacial area in a tapered bubble column

BACKGROUND: Investigations of hold up (ε_g) and interfacial area (a) in cylindrical bubble columns have been reported extensively but reported similar investigations in tapered bubble columns are sparse in the existing literature. Thus the current article reports the experimental determination of ε_g and a using a tapered bubble column. RESULTS: The present system generated ε_g (0.556 to 0.641) in a CO₂? NaOH system 20% higher than in an air–water system (0.466 to 0.534) and values were higher than in existing systems. Also, the values of ε_g in the air–water system were higher than reported for a column with shorter tapered angle. Values of ε_g fitted very well with the well‐known Akita and Yoshida correlation. The observed values of a (235 and 700 m² m^?3) were higher than values obtained (2 to 600 m² m^?3) in existing systems. The energy dissipation was 203 to 335 W m^?3, which was lower than that (100 to 1200 W/m³) in existing systems. A correlation developed to predict the pressure drop in terms of Euler number was statistically highly significant. CONCLUSION: The present research a chieved higher values of hold up and interfacial area, and lower values of energy dissipation per unit volume of dispersion compared with existing systems. Findings of the present study coupled with previous studies indicate that the tapered bubble column developed could find potential application not only in air pollution control but also in gas‐liquid mass transfer operations. Copyright © 2011 Society of Chemical Industry     

Polymerase chain reaction compatibility of adhesive transfer tape based microfluidic platforms

Laser patterned adhesive transfer tapes are a rapid, versatile, and low cost option to fabricate microfluidic platforms. In this work, we examined the compatibility with polymerase chain reaction (PCR) of different types of adhesive tape materials patterned with a CO₂ laser cutter. Acrylic, polyimide, and silicone-based tapes were considered. We performed a systematic study on off-the-shelf adhesive tapes with respect to fluid handling, PCR inhibition, reagent loss, and on-chip PCR reaction. A novel microfluidic PCR approach was implemented that combines the advantages of previously reported systems. It uses a thermal gradient from a single heating element and the thermocycling was carried out by passing the reaction mixture back and forth in a microfluidic channel strategically placed along the thermal gradient. Only the silicone-based tapes were compatible with on-chip PCR. The overall fabrication process takes less than 30 min, uses only off-the-shelf finished or semi-finished materials, and is amenable to large-scale reel-to-reel processing.     

Topic formation and development: a core-group evolving process

Recent years have witnessed increased interests in topic detection and tracking (TDT). However, existing work mainly focuses on overall trend analysis, and is not developed for understanding the evolving process of topics. To this end, this paper aims to reveal the underlying process and reasons for topic formation and development (TFD). Along this line, based on community partitioning in social networks, a core-group model is proposed to explain the dynamics and to segment topic development. This model is inspired by the cell division mechanism in biology. Furthermore, according to the division phase and interphase in the life cycle of a core group, a topic is separated into four states including birth state, extending state, saturation state and shrinkage state. In this paper, we mainly focus our studies on scientific topic formation and development using the citation network structure among scientific papers. Experimental results on two real-world data sets show that the division of a core group brings on the generation of a new scientific topic. The results also reveal that the progress of an entire scientific topic is closely correlated to the growth of a core group during its interphase. Finally, we demonstrate the effectiveness of the proposed method in several real-life scenarios.     

An ontology for unifying behavior-change literature

Changing the behavior of human operators is an underutilized approach to reduce the resource consumption of manufacturing. We created an ontology to make more accessible the existing work on behavior change, and categorized current knowledge under the headings: Problem Types, Barriers, Principles, Strategies, Mechanisms, Applications and Authors. Constructed using a web ontology language, the structure allows free navigation from any of the above category headings, and enables design practitioners better access to the strategies most relevant to their problem. We provide an example of how researchers can identify useful strategies for a specific problem in manufacturing.     

Outcoupling efficiency enhancement in organic light emitting diodes via nano-structured indium tin oxide and nano-phosphors

Nanostructures of indium tin oxide (NSITO) deposited using glancing angle deposition (GLAD) technique are used to enhance the outcoupling efficiency of blue and green OLEDs. The enhancement was found to be angle independent. The efficiency of device has increased by about 2.1 times by the use of NSITO at glass/air interface and ITO/glass interface. Similarly nano-phosphor particles were used to recover the glass modes by depositing them on the backside of the substrates. Both these methods were found to be effective for outcoupling efficiency enhancement.