Wayne Joerding

Economic growth and defense spending: Granger Causality

Abstract This paper uses Granger causality to check the assumed exogeneity of military spending relative to economic growth in previous studies of the relationship between economic growth and military spending. The test results, performed using two measures of military spending, consistently show that military spending is not a strongly exogenous variable, implying that previous work in this area is flawed.

Encoding a priori information in feedforward networks

Abstract Theoretical results and practical experience indicate that feedforward networks are very good at approximating a wide class of functional relationships. Training networks to approximate functions takes place by using exemplars to find interconnect weights that maximize some goodness of fit criterion. Given finite data sets it can be important in the training process to take advantage of any a priori information regarding the underlying functional relationship to improve the approximation and the ability of the network to generalize. This paper describes methods for incorporating a priori information of this type into feedforward networks. Two general approaches, one based upon architectural constraints and a second upon connection weight constraints form the basis of the methods presented. These two approaches can be used either alone or in combination to help solve specific training problems. Several examples covering a variety of types of a priori information, including information about curvature, interpolation points, and output layer interrelationships are presented.

Why some feedforward networks cannot learn some polynomials

It seems natural to test feedforward networks on deterministic functions. Yet, some simple functions, notably polynomials, present some difficult problems for approximation by feedforward networks. The estimated parameters become unbounded and fail to follow any unique pattern. Furthermore, as the fit to the specified functions becomes closer, numerical problems may develop in the algorithm. This paper explains why these problems occur for polynomials of order less than or equal to the number of hidden units of a feedforward network. We show that other examples occur for functions mathematically related to the networks squashing function. These difficulties do not indicate problems with the training algorithm, but occur as an inherent consequence of the role of the connection weights in feedforward networks.

Are stock prices excessively sensitive to current information

Abstract The last few years has seen publication of evidence that stock prices overreact to current information. These studies have been criticized on several grounds, that the test statistics are biased and are invalid in the presence of non-stationary dividend series. This paper develops an extension of the variance bound test which is immune to these criticisms. The paper also presents evidence that the degree of excess volatility increases as the length of time the stock is held increases. Indeed, the variance bound test is satisfied for holding periods of less than two years.

Teaching and Learning with Individually Unique Exercises

In this article, the author describes the pedagogical benefits of giving students individually unique homework exercises from an exercise template. Evidence from a test of this approach shows statistically significant improvements in subsequent exam performance by students receiving unique problems compared with students who received traditional paper assignments that were identical across students. The author also describes the software developed by himself and his students to implement this approach to homework problems. The software generates unique computer-graded assignments for each student from an assignment template and scores the resulting exercises. Instructors can create questions that require students to interact with diagrams or provide solutions to symbolic equations. The software is freely available to educators under an open-source license, to use, edit, and improve as they choose.

Global estimation of feedforward networks with a priori constraints

This paper presents a feedforward network estimation algorithm that addresses two issues, (i) avoiding local inferior minima to the performance criteria, and (ii) imposinga priori constraints to improve generalization and test economic hypotheses. The algorithm combines methods either previously developed or obviously beneficial but not yet combined. These involve combining linear least squares with simulated annealing, along with weight space reducing methods, to considerably improve its speed relative to pure simulated annealing. We present evidence on the algorithms reliability at finding a global minimum. We also demonstrate how to constrain the estimation process to find networks that satisfy a givena priori condition. We provide examples of imposinga priori information to (i) prevent the trained network from making fundamental errors and (ii) to test economically interesting hypotheses.

ISEGS: Individual Student Exercise Generating System

This paper describes software developed by the author and students to meet specific teaching goals in large economics classes. ISEGS generates unique computer-graded assignments for each student from an assignment template. ISEGS supports five types of questions: (1) Multiple-choice, (2) Interactive Diagrams, (3) Numerical, (4) Equation, (5) Short-answer. The software uses open source software (PHP, MySQL) and will run on a web server with these services. This paper also describes the current limitations of the software and experience using this software with a large introductory economics class. The software is freely available to educators, under an open source license, to use, edit, and improve as they choose.