Action-based feature representation for reverse engineering trading strategies

This paper considers the problem of reverse engineering strategies for trading in the financial markets. We investigate this problem in the context of a trading tournament in which student teams used delta hedging and other mechanisms to attempt to achieve benchmark performance in managing a hedge fund in a simulated market. Our hypothesis is that machine learning models can be trained to solve the apprenticeship learning problem; that is, these models can learn to trade like tournament participants. After reviewing classical return-matching approaches and recent work in inverse reinforcement learning, we propose a supervised learning methodology that makes use of recursive partitioning (RP). Our proposed RP approach is based on a feature representation for actions that, we argue, corresponds to the information structures readily available to tournament participants. RP achieves high accuracy in predicting the type and scale of participant trades and in tracking overall portfolio performance. Our results suggest that further research on our proposed approach is warranted and should include an expansion to testing on data from real markets.     

The maritimes region oil spill alerting and response organisation

Through cooperation among the Canadian Coast Guard, the petroleum industry, and federal and provincial environment departments, a toll free oil spill reporting telephone number (ZE 49000) was established in 1974 utilizing the facilities of the Coast Guard Traffic Centre, Maritimes. This centre advises responsible government and industry officials in accordance with established alerting procedures and all such communication regarding spillages of oil or noxious materials is recorded and retained on tape for each incident.     

Effects of ozone on inter- and intra-species competition and photosynthesis in mesocosms of Lolium perenne and Trifolium repens

Trifolium repens and Lolium perenne were exposed as both monocultures and two-species mixtures to an episodic rural ozone regime in large, well-watered containers within solardomes for 12 weeks. There were reductions in biomass for T. repens, but not L. perenne, and the proportion of T. repens decreased in ozone-exposed mixtures compared to the control. In addition, leaf biomass of T. repens was maintained at the expense of biomass partitioning to the stolons. The decreased growth corresponded with decreased photosynthetic capacity for T. repens, however, by the end of the exposure there was also decreased photosynthetic capacity of L. perenne, a species previously considered insensitive to ozone. The observed decreases in photosynthetic efficiency and capacity in elevated ozone indicate that the ability of such ubiquitous vegetation to act as a sink for atmospheric carbon may be reduced in future climates.     

Simulation of resuspended sediments resulting from dredging operations by a numerical flocculent transport model

Environmental remediations such as dredging operations cause contaminated sediments from the bottom of water bodies to become suspended into the water column. These resuspended particles are significant water quality concerns and cause adverse effects to aquatic organisms. In this paper, we present a vertically integrated two-dimensional flocculent sediment transport model to better model concentration changes of resuspended bottom sediments. The flocculent transport model has been applied to the Savannah River cutterhead dredge field study involving the resuspension of bottom sediments. The results showed that the model predictions correlate reasonably well with field data. These comparisons suggest that the flocculent sediment transport model can be used to predict the concentration profiles of a plume of toxic compounds resulting from cutterhead dredge operation.     

Molecular structures and associations of humic substances in the terrestrial environment

Here we show, for the first time, evidence of the primary molecular structures in humic substances (HS), the most abundant naturally occurring organic molecules on Earth, and their associations as mixtures in terrestrial systems. Multi-dimensional nuclear magnetic resonance (NMR) experiments show us that the major molecular structural components in the mixtures operationally defined as HS are aliphatic acids, ethers, esters and alcohols; aromatic lignin derived fragments; polysaccharides and polypeptides. By means of diffusion ordered spectroscopy, distinct diffusion coefficients consistent with relatively low molecular weight molecules were observed for all the components in the mixtures, and saccharides were the largest single class of component present. Liquid chromatography NMR confirmed that HS components can be easily separated and nuclear Overhauser effect (NOE) enhancements support the finding that the components are of relatively low molecular weight <~2,000 Da. The widely recognized properties of HS, i.e., characteristics indicative of crosslinked, macromolecular networks, can now be explained as aggregation of mixtures, most likely instigated by complexation with metal cations.     

Modelling skin permeability in risk assessment--the future

The modelling of skin permeability is important for transdermal drug delivery, in the cosmetic industry and for risk assessment attendant on dermal exposure to toxic substances. The two principal methods currently used are quantitative structure-activity relationships (QSARs), used in the main to predict permeability coefficients, and mathematical modelling based on analytical or numerical solutions to the relevant partition and transport equations and used to predict the amount of a substance permeating through the skin. This paper will assess recent progress in this area and suggest what will be needed for future advancements. The considerable effort invested in the development of QSARs during the past decade has resulted in only rather modest progress. Further significant improvement in our ability to predict percutaneous permeability is likely to require the measurement of new data under carefully controlled conditions and its fitting to new QSAR equations. Reliable assessments of risks following dermal exposures will demand new integrated mathematical models that include the variables associated with the exposure and penetration processes as well as the factors that control the subsequent passage of the penetrant into the systemic system.     

Refinement of the density-modified displacement method for efficient treatment of tetrachloroethene source zones

A novel method to remediate dense nonaqueous phase liquid (DNAPL) source zones that incorporates in situ density conversion of DNAPL via alcohol partitioning followed by displacement with a low interfacial tension (IFT) surfactant flood has been developed. Previous studies demonstrated the ability of the density-modified displacement (DMD) method to recover chlorobenzene (CB) and trichloroethene (TCE) from heterogeneous porous media without downward migration of the dissolved plume or free product. However, the extent of alcohol (n-butanol) partitioning required for in situ density conversion of high-density NAPLs, such as tetrachloroethene (PCE), could limit the utility of the DMD method. Hence, the objective of this study was to compare the efficacy of two n-butanol delivery approaches: an aqueous solution of 6% (wt) n-butanol and a surfactant-stabilized macroemulsion containing 15% (vol) n-butanol in water, to achieve density reduction of PCE-NAPL in two-dimensional (2-D) aquifer cells. Results of liquid-liquid equilibrium studies indicated that density conversion of PCE relative to water occurred at an n-butanol mole fraction of 0.56, equivalent to approximately 5 ml n-butanol per 1 ml of PCE when in equilibrium with an aqueous solution. In 2-D aquifer cell studies, density conversion of PCE was realized using both n-butanol preflood solutions, with effluent NAPL samples exhibiting density reductions ranging from 0.51 to 0.70 g/ml. Although the overall PCE mass recoveries were similar (91% and 93%) regardless of the n-butanol delivery method, the surfactant-stabilized macroemulsion preflood removed approximately 50% of the PCE mass. In addition, only 1.2 pore volumes of the macroemulsion solution were required to achieve in situ density conversion of PCE, compared to 6.4 pore volumes of the 6% (wt) n-butanol solution. These findings demonstrate that use of the DMD method with a surfactant-stabilized macroemulsion containing n-butanol holds promise as an effective source zone remediation technology, allowing for efficient recovery of PCE-DNAPL while mitigating downward migration of the dissolved plume and free product.     

A Technique for Plume Visualization in Power Plant Siting

Because of the increasing demand for energy, construction of a number of large power plants is planned, particularly in the southwestern U.S. Before construction of a plant can begin, however, it must be shown that its emissions will not degrade air quality beyond certain acceptable limits. A new plant must undergo not only a New Source Review¹ to demonstrate compliance with New Source Performance Standards,² but also an analysis to show that it will not produce significant deterioration of air quality in designated pristine areas (often these areas are state or federal parks.) Further, as a result of the 1977 amendments³ to the Clean Air Act, visibility must not be impaired where impairment is defined as a reduction in visual range or as atmospheric discoloration. Plume visual impact must thus be considered before permission for construction will be granted. Because of these requirements, it is important to be able to visualize the plume from a power plant as it travels through surrounding terrain.