No abstract

Transfer learning aims at improving the performance of target learners on target domains by transferring the knowledge contained in different but related source domains. In this way, the dependence on a large number of target domain data can be reduced for constructing target learners. Due to the wide application prospects, transfer learning has become a popular and promising area in machine learning. Although there are already some valuable and impressive surveys on transfer learning, these surveys introduce approaches in a relatively isolated way and lack the recent advances in transfer learning. As the rapid expansion of the transfer learning area, it is both necessary and challenging to comprehensively review the relevant studies. This survey attempts to connect and systematize the existing transfer learning researches, as well as to summarize and interpret the mechanisms and the strategies of transfer learning in a comprehensive way, which may help readers have a better understanding of the current research status and ideas. Different from previous surveys, this survey paper reviews over forty representative transfer learning approaches from the perspectives of data and model. The applications of transfer learning are also briefly introduced. In order to show the performance of different transfer learning models, twenty representative transfer learning models are used for experiments. The models are performed on three different datasets, i.e., Amazon Reviews, Reuters-21578, and Office-31. And the experimental results demonstrate the importance of selecting appropriate transfer learning models for different applications in practice.

An analysis of the spatiotemporal variability in summer precipitation during the period 1961-2010 is presented based on monthly precipitation datasets from 66 meteorological stations in the central and eastern Tibetan Plateau (TP). By applying empirical orthogonal function (EOF) analysis, a strong reversal is found in the variability of summer precipitation between the northeastern and the southeastern TP on the inter-annual timescale; this reversal is defined as the Dipole Oscillation in summer precipitation over the TP.Our analysis shows that the North Atlantic Oscillation (NAO) greatly controls the Dipole Oscillation in TP summer precipitation by modifying the atmospheric circulation over and around the TP. With the increased stationary wave activity spreading eastward from the North Atlantic to the TP, a pronounced wave train pattern bridges the North Atlantic Ocean and the TP. During the positive phase of the NAO, warm moist air from the oceans around Asia is transported by the southeastern flank of the anticyclonic anomaly over East Asia to the northeastern TP. This northward-moving warm moist air encounters cold air masses transported by the northwestern flank of the cyclonic anomaly over the northeastern TP. The confluence of the cold and warm air masses subsequently strengthens cumulus convective activities and ultimately results in excessive precipitation over the northeastern TP. Meanwhile, as a cyclonic anomaly sets up over northwestern India and Pakistan, water vapour condenses into precipitation over northwestern India and Pakistan, inhibiting Arabian Sea moisture inflows into the southeastern TP and northeastern India. As a result, a precipitation deficit occurs over the southeastern TP. The opposite scenario occurs during the negative phase of the NAO.

[1] The influence of anthropogenic pollution on the region of Tianshan Mountain, a remote area in arid central Asia, has been debated in the recent years. An ice core, covering the past 43 years, retrieved from Glacier 1 at Urumqi River head in the east Tianshan, northwest China, was analyzed to examine the problem. , the major inorganic anion in the core, averages 232.9 ± 279.9 ng g À1 (N = 542). The organic and inorganic records have covaried in the past four decades. They originate principally from anthropogenic pollution, coal combustion in particular, of the local and regional atmosphere. The pH values in the record range from 6 to 9 with an average of 6.9 ± 0.5 (N = 541). The general trend of the pH data matches that of HCOO , indicating that the anthropogenic pollution has released considerable particulate material along with unsaturated hydrocarbons and SO 2 . As a result, the pollution has not been acidifying the environment, but making it alkaline.

No abstract

A survey of July 1st glacier, Qilian Shan, China, was carried out in 2002. Previously, the glacier’s boundary had been recorded in 1956, and further research had been carried out in the mid- 1970s and 1980s. Our survey reveals that area shrinkage and surface lowering have accelerated in the past 15 years. Surface elevation changes can result from changes in accumulation, surface melting and emergence velocity. The contributions of these elements to surface lowering are evaluated at the lower part of the glacier from observations of surface velocity, ice thickness and precipitation, and from temperature data near the glacier. Apart from the effect of glacier ice redistribution, our analysis reveals quantitatively that the recent accelerated glacier shrinkage has been caused by increasing temperature. Furthermore, it is established that meltwater discharge from the glacier in the past 17 years has increased due to glacier shrinkage, by about 50% over that from 1975 to 1985.

No abstract

No abstract