科研成果

Forecast combinations have been widely applied in the last few decades to improve forecasting. Estimating optimal weights that can outperform simple averages is not always an easy task. In recent years, the idea of using time series features for forecast combinations has flourished. Although this idea has been proved to be beneficial in several forecasting competitions, it may not be practical in many situations. For example, the task of selecting appropriate features to build forecasting models is often challenging. Even if there was an acceptable way to define the features, existing features are estimated based on the historical patterns, which are likely to change in the future. Other times, the estimation of the features is infeasible due to limited historical data. In this work, we suggest a change of focus from the historical data to the produced forecasts to extract features. We use out-of-sample forecasts to obtain weights for forecast combinations by amplifying the diversity of the pool of methods being combined. A rich set of time series is used to evaluate the performance of the proposed method. Experimental results show that our diversity-based forecast combination framework not only simplifies the modeling process but also achieves superior forecasting performance in terms of both point forecasts and prediction intervals. The value of our proposition lies on its simplicity, transparency, and computational efficiency, elements that are important from both an optimization and a decision analysis perspective.

Forecasting has always been at the forefront of decision making and planning. The uncertainty that surrounds the future is both exciting and challenging, with individuals and organisations seeking to minimise risks and maximise utilities. The large number of forecasting applications calls for a diverse set of forecasting methods to tackle real-life challenges. This article provides a non-systematic review of the theory and the practice of forecasting. We provide an overview of a wide range of theoretical, state-of-the-art models, methods, principles, and approaches to prepare, produce, organise, and evaluate forecasts. We then demonstrate how such theoretical concepts are applied in a variety of real-life contexts. We do not claim that this review is an exhaustive list of methods and applications. However, we wish that our encyclopedic presentation will offer a point of reference for the rich work that has been undertaken over the last decades, with some key insights for the future of forecasting theory and practice. Given its encyclopedic nature, the intended mode of reading is non-linear. We offer cross-references to allow the readers to navigate through the various topics. We complement the theoretical concepts and applications covered by large lists of free or open-source software implementations and publicly-available databases.

Bacteria present in natural environment especially those in cold regions would experience freeze-thaw (FT) process during day-night and season turns. However, knowledge about the influence of FT on bacteria release behaviors in porous media was limited. In present study, the bacteria release behaviors from quartz sand columns without and with 1 and 3 FT treatment cycles under three water saturations (θ=100%, 90%, and 60%) were investigated. We found that for all three water saturated columns without FT treatment, negligible bacteria released from columns via background salt solution elution, while the subsequent release of bacteria from sand columns via low ionic strength (IS) solution elution decreased with decreasing column water saturations. More importantly, we found unlike the negligible bacteria release in columns without FT treatment, for columns with high saturations (θ=100% and 90%), FT treatment could promote bacteria release with background salt solution elution. Moreover, for high saturated columns, FT treatment would decrease subsequent bacteria release with low IS solution elution. This phenomenon was more obvious with increasing FT treatment cycles. In contrast, FT treatment had negligible influence on bacteria release from columns with lower saturation (θ=60%). The decreased bacterial sizes, the loss of bacterial flagella, as well as the change of local configuration of porous media (via changing water into ice and ice back into water) during the FT processes contributed to increased bacteria release via background salt solution elution from high saturated sand columns. While, the reduced amount of bacteria being retained at secondary energy minima drove to the subsequently decreased bacteria release via low IS solution elution. The results of this study clearly showed that for porous media with high saturations, FT cycles would increase the risk of bacteria detaching from porous media with flushing by the background solution.

Strontium isotope composition is a powerful proxy for tracking the source and evolution of a range of geological processes in magmatic, sedimentary and metamorphic environments. Owing to the mobility of Sr with fluids, Sr isotope is also an ideal proxy to investigate, and has been extensively used to tackle the history of magmatic-hydrothermal systems and associated water-rock interactions. Apatite is an excellent choice for such a purpose as it is ubiquitous in almost all these environments, and contains abundant Sr but exceptionally limited Rb. Microanalytical techniques are preferred for apatite Sr isotope studies since most apatites are small in size and may have a complicated growth history to be decoded. Matrix-matched reference materials are a prerequisite for microanalysis, and although variable F- and OH-rich apatite Sr isotope reference materials have been developed, chlorine-rich ones are lacking. Here, we further evaluate Eppawala-AP, a Cl-rich apatite, as a potential Sr isotope reference material. Our extensive Sr isotope microanalysis at three laboratories confirmed that the Eppawala-AP was homogeneous in terms of Sr isotope composition, with a recommended Sr-87/Sr-86 ratio of 0.704989 +/- 0.000017 (2 SD, n = 9) determined by thermal ionization mass spectrometry from another two laboratories. The Eppawala-AP also had homogeneous trace element contents and Nd isotope composition. Therefore, Eppawala-AP can be used as a Sr isotope reference material, and also as a working reference to monitor the Nd isotope and trace element analysis, and thus could be expected to fill the much-needed gap for Cl-rich apatite reference materials and facilitate more precise and accurate analysis.

This chapter investigates China's recent higher education reform by developing applied education to addressing the issue of underemployment and unemployment of college graduates. Reviewing China's highly segmented labor market and the growth of flexible employment, we suggest that the negative effect of overeducation is mainly on those college graduates employed in the informal sector. Our analysis identifies a restructuring process of HE system aiming to break down the homogeneity among the elite and non-elite universities by making a dual process of differentiation. We conclude that it is difficult to achieve the ``parity of esteem'' between academic and applied education within a steeply stratified HE system and suggest that the relationship between HE system and cooperate employers deserves an in-depth investigation to turn around the fortunes of graduates from lower-tier universities.