Advances in supercomputers have come at a steady pace over the past 20 years. The next milestone is to build an Exascale computer however this requires not only speed improvement but also significant enhancements for energy efficiency and massive parallelism. This paper examines technological progress of supercomputer development to identify the innovative potential of three leading technology paths toward Exascale development: hybrid system, multicore system and manycore system. Performance measurement and rate of change calculation were made by Technology Forecasting using Data Envelopment Analysis (TFDEA.) The results indicate that the current level of technology and rate of progress can achieve Exascale performance between early 2021 and late 2022 as either hybrid systems or manycore systems.

The Toyota Prius was first introduced in 1997 and since then over 150 hybrid electric vehicles (HEVs) have been brought to the automobile market around the world. This was spurred by a major interest in the future of vehicles using 'alternative fuel' for addressing environmental and fuel dependency concerns. Based upon previous work 1 , which identified an input-output model that could successfully explain the progress of HEV technologies, this study evaluates and compares the technological advancement observed in different HEV market segments over the past 15 years. The results indicate that the introduction of a wide range of midsize HEVs is posing a threat to the two-seaters and compact HEV segments while an SUV segment shows a fast adoption with a significant performance improvement. The rates of change for each segment are also provided to give insights into the estimation of the future performance levels for new product development target setting purposes.

Imbalanced classes are a common problem in machine learning, and the computational costs required for proper resampling increases with the data size. In this study, a simple and effective undersampling method, named particle stacking undersampling (PSU) was proposed. Compared with other competing undersampling methods, PSU can significantly reduce the computational costs, while minimizing information loss to prevent a prediction bias. The performance benchmark applied on 55 binary classification problems indicated that the proposed method not only achieved an enhanced classification performance over other well-known undersampling methods (random undersampling, NearMiss-1, NearMiss-2, cluster centroid, edited nearest neighbor, condensed nearest neighbor, and Tomek Links) but also provided a computational simplicity that can be scalable to large data. Moreover, an experiment verified that two propositions forming the basis of the PSU algorithm can also be applied to other undersampling methods to achieve methodological improvements. INDEX TERMS Data mining, imbalanced data, undersampling, big data, support vector machines.

This chapter illustrates the Technology Forecasting using Data Envelopment Analysis (TFDEA) process on Liquid Crystal Display (LCD) performance characteristics from 1997 to 2012. The objective of this study is to forecast future state-of-the-arts (SOAs) specifications as well as to diagnose past technological advancement of the LCD industry. Appropriate characteristics were determined from a group of LCD technologists. Data was gathered from public databases and outlying data points were cross-referenced as a validity check. The TFDEA process is defined and its application to the dataset is described in detail. The results not only provide information on how LCD industry has evolved but also provide an insight on future NPD targets.

In this paper, we propose a technology trajectory mapping approach using data envelopment analysis (DEA) that can scrutinize technology progress patterns from multidimensional perspectives. Literature reviews on technology trajectory mappings have revealed that it is imperative to identify key performance measures that can represent different value propositions and then apply them to the investigation of technology systems in order to capture indications of the future disruption. The proposed approach provides a flexibility not only to take multiple characteristics of technology systems into account, but also to deal with various trade-offs among technology attributes by imposing weight restrictions in the DEA model. The application of this approach to the flat panel technologies is provided to give a strategic insight for the players involved.