Comparison of machine learning approaches and statistical methods for estimation of airborne chemical exposures

Accurate quantitative estimation of exposure using retrospective data has been one of the most challenging tasks in the exposure assessment field. To improve these estimates, some models have been developed using published exposure databases with their corresponding exposure determinants. These models are designed to be applied to reported exposure determinants obtained from study subjects or exposure levels assigned by an industrial hygienist, so quantitative exposure estimates can be obtained. ^ In an effort to improve the prediction accuracy and generalizability of these models, and taking into account that the limitations encountered in previous studies might be due to limitations in the applicability of traditional statistical methods and concepts, the use of computer science- derived data analysis methods, predominantly machine learning approaches, were proposed and explored in this study. ^ The goal of this study was to develop a set of models using decision trees/ensemble and neural networks methods to predict occupational outcomes based on literature-derived databases, and compare, using cross-validation and data splitting techniques, the resulting prediction capacity to that of traditional regression models. Two cases were addressed: the categorical case, where the exposure level was measured as an exposure rating following the American Industrial Hygiene Association guidelines and the continuous case, where the result of the exposure is expressed as a concentration value. Previously developed literature-based exposure databases for 1,1,1 trichloroethane, methylene dichloride and, trichloroethylene were used. ^ When compared to regression estimations, results showed better accuracy of decision trees/ensemble techniques for the categorical case while neural networks were better for estimation of continuous exposure values. Overrepresentation of classes and overfitting were the main causes for poor neural network performance and accuracy. Estimations based on literature-based databases using machine learning techniques might provide an advantage when they are applied to other methodologies that combine `expert inputs' with current exposure measurements, like the Bayesian Decision Analysis tool. The use of machine learning techniques to more accurately estimate exposures from literature-based exposure databases might represent the starting point for the independence from the expert judgment.^

An analysis of the revenue policy -making process of the Texas Legislature

The purpose of this dissertation was to develop a conceptual framework which can be used to account for policy decisions made by the House Ways and Means Committee (HW&MC) of the Texas House of Representatives. This analysis will examine the actions of the committee over a ten-year period with the goal of explaining and predicting the success of failure of certain efforts to raise revenue.^ The basis framework for modelling the revenue decision-making process includes three major components--the decision alternatives, the external factors and two competing contingency theories. The decision alternatives encompass the particular options available to increase tax revenue. The options were classified as non-innovative or innovative. The non-innovative options included the sales, franchise, property and severance taxes. The innovative options were principally the personal and corporate income taxes.^ The external factors included political and economic constraints that affected the actions of the HW&MC. Several key political constraints on committee decision-making were addressed--including public attitudes, interest groups, political party strength and tradition and precedents. The economic constraints that affected revenue decisions included court mandates, federal mandates and the fiscal condition of the nation and the state.^ The third component of the revenue decision-making framework included two alternative contingency theories. The first alternative theory postulated that the committee structure, including the individual member roles and the overall committee style, resulted in distinctive revenue decisions. This theory will be favored if evidence points to the committee acting autonomously with less concern for the policies of the Speaker of the House. The Speaker assignment theory, postulated that the assignment of committee members shaped or changed the course of committee decision-making. This theory will be favored if there was evidence that the committee was strictly a vehicle for the Speaker to institute his preferred tax policies.^ The ultimate goal of this analysis is to develop an explanation for legislative decision-making about tax policy. This explanation will be based on the linkages across various tax options, political and economic constraints, member roles and committee style and the patterns of committee assignment. ^

The dynamic evaluation of the learning potential: a procedure

Fil: Malbrán, María del Carmen. Universidad Nacional de La Plata. Facultad de Humanidades y Ciencias de la Educación; Argentina.

The dynamic evaluation of the learning potential: a procedure

Fil: Malbrán, María del Carmen. Universidad Nacional de La Plata. Facultad de Humanidades y Ciencias de la Educación; Argentina.

Sequential variation of carbonate cycles in South Atlantic DSDP locations

We report well-dated Late Cretaceous and Early Tertiary precessional climatic cycles, recorded by rhythmic carbonate maxima and minima in South Atlantic deep sea sites. Spectral analyses of digitized sediment color, a suitable carbonate proxy, show prominent regularities in the spacing marl-carbonate beds. Magnetostratigraphic dating over a number of magnetic chrons constrains the duration of the cycles, which can be detected over at least 20 Myr of sedimentation at 7 coring locations. Their mean absolute period of 23.5 +/- 4.4kyr agrees closely with the predicted late Cretaceous precessional period of 20.8 kyr. Because they can be matched to a physical forcing mechanism with a known repeat time, the cycles offer a new high-resolution tool to measure rates of climate change before and after the Cretaceous-Tertiary (K/T) boundary. From counts of carbonate cycles, we derive the position of the K/T boundary within C29R at 350 kyr after the base of the reversal. The constancy of cycle thickness (linearly related to sedimentation rate) and amplitude up to the "boundary clay" does not give evidence for climate instability preceding the boundary. Orbital chronometry records a step-function decrease in sediment accumulation rate at the Cretaceous-Tertiary boundary that is consistent with a geologically instantaneous event.