Search

Lung cancer is one of the most fatal disease with high lethality. In general lung cancers are diagnosed by radiologists. But checking radiological image is a very toilsome work for radiologists because it requires long time practice and high concentration. So, many computer-aided diagnosis (CAD) systems were introduced to cooperate with radiologists and nowadays lots of CAD systems based upon deep learning exceed human experts in diagnosing accuracy. And the remarkable thing is that the much of progress has been made in designing architectures. But, in this paper, a new pre-processing method (lung-range-standardization) is proposed in order to improve the general accuracy of lung-related diagnosis systems and to increase the utility of LIDC dataset. And the efficiency of the proposed pre-processing method is validated through comparison between the nodule segmentation model trained using lung-range-standardization and the nodule segmentation model, which is trained without lung-range-standardization.

Separation of NaCN from the mixture of NaCN and Na2CO3 (which is mainly produced by urea and sodium carbonate method) was studied. The soluble possibility of NaCN and Na2CO3 in different mixed solvents was theoretically investigated. The experimental solubility of NaCN and Na2CO3 in water, ethanol and mixed solvent (water + ethanol) at the temperature range of 285.15K-323.15K was determined by a static method. e-NRTL model was used to estimate the experimental solubility of NaCN and Na2CO3. It is found out that optimal separation of NaCN ranges at the temperature of 318.15K below and at the ethanol mass composition of 0.7–0.8.

The solubility of NaCN and Na2CO3 in the water–ethanol mixture was determined through experiments. Also the polynomial fit was used to lead to the relationship between the solubility of NaCN/Na2CO3 and the solvent composition. In the meantime, the solvent molecular-electrolyte pair interaction parameters of electrolyte NRTL (eNRTL) model were estimated by using the data regression function of Aspen Plus software from solubility data of NaCN and Na2CO3 in water–ethanol mixed solvent. The process simulation for separation and purification of NaCN and Na2CO3 in water–ethanol mixed solvent was carried out by using Aspen Plus software. During the simulation, electrolyte equilibrium system in water–ethanol mixed solvent was observed by using eNRTL model property method, and the influence of several factors on the separation efficiency of NaCN was examined.

Although extreme learning machine (ELM) provides better generalization performance at a much faster learning speed than traditional learning algorithms, the classical ELM can not obtain ideal results for the imbalanced data problem. In this paper, in order to conquer the learning capability of the classical ELM for an imbalance data learning, we define a new cost function of ELM optimization problem based on G-mean widely used as evaluation metric in imbalance data learning. We perform experiments on standard classification datasets which consist of 58 binary datasets and 11 multi-class datasets with the different degrees of the imbalance ratio. Experimental results show that proposed algorithm can improve the classification performance significantly compared with other state-of-the-art methods. We also demonstrate that our proposed algorithm can achieve high accuracy in representation learning by performing experiments on YouTube-8M with feature representation from convolutional neural networks. Statistical results indicate that the proposed approach not only outperforms the classical ELM, but also yields better or at least competitive results compared with several state-of-the-art class imbalance learning approaches.