High Runtime Research Articles

The complexity of a simple stochastic OR-tree model in which “directional search” is bad

A simple stochastic model of OR-trees is investigated in which the directional search algorithm has a very high average run time. The optimal algorithms just make the opposite of directional search: they spread their questions over the tree as balanced as possible. This is the first time that in a nontrivial stochastic recursion tree model the optimality of a nondirectional algorithm with respect to the average run time has been proved.

CL: a flexible and efficient tool for constructing knowledge-based expert systems

A description is given of CL, the conceptual network-based language, which is a Lisp-based tool kit to assist knowledge engineers in customizing a general framework into an optimal one for a specific domain and achieving high runtime performance with user-defined control strategies. The design objectives embodied in CL are discussed, and its extensibility and domain language primitives are examined. It is shown how the use of CL can improve runtime performance. Experience with CL and its current status are described. >

Dispersion Modelling for a Complex Estuary – The Case of the Tagus

Estuarine dispersion models have been commonly used to define the pollutant loads permissible to achieve pre-defined water quality levels and improve our knowledge of estuarine phenomena. Those models for large estuaries with complex hydrodynamic and ecological processes usually have extremely high running times. This paper presents an approach based on the use of increasingly complex models, which attempts to circumvent the problem of initial lack of data, as well as to give some initial insight into the processes of the Tagus Estuary, within acceptable levels of precision. As a first stage, simple models were developed and applied to the estuary, one of the largest in Europe, with more than 300 sources of pollution and intensive use for recreation, fishing, and navigation. The computational exercises undertaken with these models were also used to accumulate information on the response of the Tagus Estuary to a number of forcing conditions. This information, synthesized in ‘if … then' rules, was integrated to form a data base on the estuary, which is currently being developed. The data base will organize existing information and, providing that learning mechanisms are included, it will also create new knowledge, as well as supplying the complex models under development with reasonable initial values.

Adaptation of the equipment for high-performance electrophoresis to isoelectric focusing

Fast and reproducible isoelectric focusing experiments have been performed in glass capillaries of 0.2 mm I.D., wall thickness 0.1 mm, and length 120 mm. These narrow, thin-walled tubes permit rapid removal of the Joule heat to allow high voltages (3000 V) and short run times. The experiments were carried out in equipment designed for high-performance electrophoresis. The protein zones were detected by on- or off-tube measurements of their UV absorption. This detection technique requires that the proteins can be mobilized following the isoelectric focusing step. This mobilization has been achieved either by pumping the zones past the stationary UV monitor (with the voltage applied to avoid zonal broadening during the elution) or by electrophoretic elution, accomplished by replacing the acid at the anode by a base (or the base at the cathode by an acid), which caused the pH gradient— and thereby the proteins— to move in the tube. The electrophoretic mobilization procedure is more universal because it is also applicable when the focusing is performed in a gel

High-performance electrophoresis: the electrophoretic counterpart of high-performance liquid chromatography

High-performance electrophoresis is the electrophoretic equivalent of high-performance liquid chromatography. Very efficient cooling is obtained because the experiments are performed in glass tubes of small diameter (0.05–0.3 mm) with a wall thickness of only 0.1–0.2 mm. Under these conditions the thermal deformation of a zone is small even at high field strengths, which allows high resolution and short run times. The solutes are recorded directly in the glass tube as they migrate electro-phoretically through the stationary UV detector of a free-zone electrophoresis or gel-scanning apparatus. Accordingly there is no delay in the detection caused by time-consuming derivatization (staining). The electrophoresis can be performed both in sieving and non-sieving gels and also in carrier-free solution. An analysis, which may comprise low-molecular-weight compounds, proteins, nucleic acids, viruses, etc., takes at present 10–60 min. This time can be considerably reduced by using a liquid-cooled electrophoresis tube. About 0.01–1 μg of material is suitable in the case of proteins.

Effect of previous exposure on response decrement induced by an olfactory stimulus.

Two experiments examined the influence of pre-exposure on the decrement produced in a trained running response by a novel odor. In Exp. 1, when first presented in the runway, the smell led to higher running times in a group of rats pre-exposed 72 hr. earlier and in a group with no pre-exposure than in a group pre-exposed 25 min. earlier. There was no significant difference between the first two groups. In Exp. 2, three groups of rats were pre-exposed, 25 min. before testing in the runway, in pre-exposure boxes which differed in degree of similarity to the runway. Their performance did not differ significantly on test sessions. Taken together, the results indicate that contact with a novel odor is sufficient to lessen its decremental effect even if it occurs outside the training situation and is not followed by reinforcement. The effects of pre-exposure dissipate with time and do not seem to depend strongly on contextual stimuli.