Search

Semi-supervised learning aims to utilize both labelled and unlabelled data to improve learning performance. This paper shows a distinct way to exploit unlabelled data for traditional semi-supervised learning methods, such as self-training. Self-training is a well-known semi-supervised learning algorithm which iteratively trains a classifier by bootstrapping from unlabelled data. Standard self-training barely selects unlabelled examples for training set augmentation according to the current classifier model, which is trained only on the labelled data. This could be problematic since the underlying classifier is not strong enough, especially when initial labelled data is sparse. Consequently, self-training suffers from too much classification noise accumulated in the training set. In this paper, we propose a novel self-training style algorithm, which exploits a manifold assumption to optimize the self-labelling process. Unlike standard self-training, our algorithm utilizes labelled and unlabelled data as a whole to label and select unlabelled examples for training set augmentation. In detail, two measures are employed to minimize the effect of noise introduced to the labelled training set: a transductive method based on controlled graph random walk is incorporated to generate reliable predictions on unlabelled data; secondly, the mechanism is adopted to sequentially augment the training set. Empirical results suggest that the proposed method can effectively improve classification performance.

Silica/titania (SiO2/TiO2) (Si/Ti= 1:1) composite aerogels have been manufactured via ambient pressure drying using tetraethoxysilane (TEOS) and tetrabutyl titanate (TBT) as precursors. Changes in the structure and the surface properties of the aerogels during the heat treatment were investigated in detail by N-2 adsorption-desorption techniques, SEM, TEM, XRD, FTIR, and UV-vis absorption spectra. The results showed that the as obtained aerogel with a porous structure possessed a surface area of 667 m(2)/g, average pore size 3.4 nm, and pore volume 0.68 cm(3)/g. After calcination, the surface area decreased while the average pore size and the pore volume increased with the rise of heat treatment temperature. FUR spectra showed the existence of Si-O-Ti in the aerogels and that after calcination at 800 degrees C, anatase nanoparticle appeared as demonstrated in XRD pattern. The photocatalytic performance was evaluated by degradation of MB and the results demonstrated that the aerogels calcined at 800 degrees C have a better photocatalytic activity than other samples. (C) 2014 Elsevier Ltd and Techna Group S.r.l. All rights reserved.

Cool coatings (solar reflective coatings) possess a high total solar reflectance and a high thermal emittance, and they have potential applications in building surfaces, pavements, and oil tanks. By decreasing the accumulation of heat, these coatings can reduce the urban heat island effect to increase thermal comfort and to decrease energy consumption in hot zones. In this work, a cool white coating based on waterborne acrylic was prepared. The reflectance and thermal emittance were measured using an ultraviolet-visible-near-infrared spectrophotometer and the steady-state calorimetric technique, respectively. The cool coating possessed a total solar reflectance of 0.887 and a thermal emittance of 0.85. The linear fitting results indicated that the thickness of the dry film is an important factor that affects the reflectance and that the physical limit of total solar reflectance in practical cool coatings is 0.91. A homemade device developed according to military specification MIL-E-46096C was used to measure insulation temperature differences, and an insulation temperature difference of 29 °C was obtained. The above results demonstrated that the cool coating can be applied to buildings.

This paper shows how to use labeled and unlabeled data to improve inductive models with the help of transductivemodels.We proposed a solution for the self-training scenario. Self- training is an effective semi-supervised wrapper method which can generalize any type of supervised inductive model to the semi-supervised settings. it iteratively refines a inductive model by bootstrap from unlabeled data. Standard self-training uses the classifier model(trained on labeled examples) to label and select candidates from the unlabeled training set, which may be problematic since the initial classifier may not be able to provide highly confident predictions as labeled training data is always rare. As a result, it could always suffer from introducing too much wrongly labeled candidates to the labeled training set, which may severely degrades performance. To tackle this problem, we propose a novel self-training style algorithm which incorporate a graph-based transductive model in the self-labeling process. Unlike standard self-training, our algorithm utilizes labeled and unlabeled data as a whole to label and select unlabeled examples for training set augmentation. A robust transductive model based on graph markov random walk is proposed, which exploits manifold assumption to output reliable predictions on unlabeled data using noisy labeled examples. The proposed algorithm can greatly minimize the risk of performance degradation due to accumulated noise in the training set. Experiments show that the proposed algorithm can effectively utilize unlabeled data to improve classification performance.

In this paper, a novel inductive support vector machine for semi-supervised learning, named IS3VM, is proposed, which aims to improve SVM by bootstrapping unlabeled data with self-training. The SVM classifier is iteratively refined through the augmentation of the training set. An improved self-training method is given by employing neighborhood graph for guarantying the reliability of newly added training examples. In detail, in each iteration of the self-training process, the local cut edge weight statistic is used to help estimate whether a newly labeled example is reliable or not, and only the reliable self-labeled examples are used to enlarge the labeled training set. Experiments show that, the improved self-training is beneficial and the proposed IS3VM algorithm can effectively exploit unlabeled data to achieve better performance, and is comparable to the-state-of-the-art semi-supervised SVM.