Deterministic Weight Research Articles

Deterministic weight evolution algorithm for solvinglocalminima problem

A new algorithm is proposed to enable local minima to be escaped from by changing the weights of a multi-layer neural network in a deterministic way. Simulation results show that the new algorithm always outperforms other traditional methods in achieving global convergence.

How important is sound domestic macroeconomics in attracting capital inflows to developing countries?

Many developing countries face deficiencies of domestic saving and foreign exchange. With foreign aid declining in real terms it becomes increasingly important to ask what they can do to attract private capital. Without capital inflows, shortages of external financing are likely to constitute an effective constraint on economic development. Within this context a key question is ‘how important is it to get the macroeconomics right?’. Can developing countries expect to be rewarded for improved macroeconomic performance by capital inflows? Clearly if they can, there is an additional incentive to seek such improvement. The theoretical and empirical analysis in this paper suggests that there will be no short-run pay-off to improved macroeconomics, beyond a point at which severe macroeconomic disequilibria are eliminated. Whereas there are predictable penalties for getting the macroeconomics badly wrong, there are no equivalently predictable rewards getting it ‘right’. In large measure this is the consequence of the ambiguities surrounding what is sound macroeconomics. Macroeconomics is simply too uncertain to encourage investors to attach deterministic weights to indicators of macroeconomic policy and performance. For as long as these uncertainties remain, it is difficult to see how domestic macroeconomics will become a dominant factor in explaining capital flows. Copyright © 1999 John Wiley & Sons, Ltd.

Estimation of the relative risk following group sequential procedure based upon the weighted log-rank statistic

In this paper we consider a group sequentially monitored trial on a survival endpoint, monitored using a weighted log-rank (WLR) statistic with deterministic weight function. We introduce a summary statistic in the form of a weighted average logged relative risk and show that if there is no sign change in the instantaneous logged relative risk, there always exists a bijection between the WLR statistic and the weighted average logged relative risk. We show that this bijection can be consistently estimated at each analysis under a suitable shape assumption, for which we have listed two possibilities. We indicate how to derive a design-adjusted p-value and confidence interval and suggest how to apply the bias-correction method. Finally, we document several decisions made in the design of the NLST interim analysis plan and in reporting its results on the primary endpoint.