Improving the accuracy of decision tree induction by feature preselection

Selecting the right set of features for classification is one of the most important problems in designing a good classifier. Decision tree induction algorithms such as C4.5 have incorporated in their learning phase an automatic feature selection strategy, while some other statistical classification algorithms require the feature subset to be selected in a preprocessing phase. It is well known that correlated and irrelevant features may degrade the performance of the C4.5 algorithm. In our study, we evaluated the influence of feature preselection on the prediction accuracy of C4.5 using a real-world data set. We observed that accuracy of the C4.5 classifier could be improved with an appropriate feature preselection phase for the learning algorithm. Beyond that, the number of features used for classification can be reduced, which is important for image interpretation tasks since feature calculation is a time-consuming process.     

Semicontinuous Discharge of Non‐Newtonian Filter Cakes in Electrofiltration

A novel mode of operation for electrofiltration is presented. The newly developed operation procedure allows for cake removal without disassembling the chamber. Highly viscous filter cakes are discharged from the filtration chamber by inserting air which permits a rapid restart of the process. The cleaning time could be reduced from about 10 min to a few seconds. The semicontinuous operation is liable to full automation and allows for cake discharge and filtration in a short period of time. In successive cycles was proved that the process efficiency is maintained using the same membrane. A fully continuous system can be accomplished by assembling several units in parallel in a single filtration press.     

Intentionality and knowledge in children's judgments of actor's responsibility and recipient's emotional reaction.

Two conflicting developmental accounts of how mental states are used in evaluating actors are tested by varying actors' intentionality, foreknowledge of outcome, and the values of motive and outcome. In Experiment 1, children judged a recipient's emotional reaction to three types of event: intended outcome, foreseen accident, and unforseen accident. Both 6- and 7-year-olds used intentionality and knowledge in their judgments of good and bad outcomes. Three-year-olds did not distinguish between accidents differing in actors' foreknowledge, but discriminated between intended and accidental outcomes when the accident was unforseen. In Experiment 2, children judged actor's responsibility for accidentally caused bad outcomes. Seven-year-olds, but not 5-year-olds, blamed actors for foreseen accidents more than for unforeseen accidents regardless of motive value. The results suggest that children use intentionality before knowledge in judgments of action sequences, and that actor's foreknowledge of an outcome influences children's ability to judge the intended/accidental distinction. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Ozone Levels in Central Israel

Measurements performed during the early summer months of 1988–1991 at a rural site in central Israel, some 50 km east of the highly urbanized coastal region, have shown that during the afternoon hours the area was often under the influence of ozone mixing ratios above the Israel ambient standard (117 ppbv) and occasionally even above 150 ppbv. Analysis of air mass back trajectories has shown that only those air masses passing over the Tel Aviv metropolitan area cause elevated ozone mixing ratios at the rural site. This highly urbanized region emits large amounts of precursors which are entrapped in the air parcels entering Israel under the predominantly westerly wind flows. As these air masses travel inland, sufficient time is available (3–5 h) to allow the photochemical reactions to generate ozone before reaching the rural site. The above hypothesis is further supported by the fact that parallel to the increase of ozone at the rural site, elevated carbon monoxide (up to 0.8 ppmv) and other trace gases were also observed. A significant correlation (R² > 0.8) was found to exist between the ozone mixing ratio and the NO_x concentrations in photochemically aged air masses. In several cases an excess of up to 12 ozone molecules was formed for each NO_x molecule present.     

Online Detection of Ellipsoidal Bubbles by an Innovative Optical Approach

Bubble and droplet size distribution influence chemical processes markedly. In order to enable an in‐line analysis of these distributions, a new single‐sided endoscopic version of the Optical Multimode Online Probe (OMOP) was developed. The acquired images are analyzed to gain bubble size distributions (BSDs) and main diagonals. A specific workflow was developed to detect and analyze overlapping bubbles based on standard image processing techniques. It can handle overlapping bubbles of both circular and elliptical shapes. Results confirmed that the combination of OMOP and the image processing workflow leads to robust detection of bubbles from two‐dimensional images. The new single‐sided OMOP proved to be a suitable tool for the acquisition of a BSD at industrial scale within a reasonable computing time.     

An integrated simulation model for European electricity and natural gas supply

Contents This paper describes an empirical long-run simulation model for the European electricity and natural gas market. In a first step, electricity and gas markets are modelled separately by dynamic linear programming. In a second step, the models are linked by iteration. The model results show significant interrelationships between gas consumption in power generation and supply conditions on the gas market. Received: 30 April 2001 / Accepted: 7 June 2001