Identifying the informative genes has always been a major step in microarray data analysis. The complexity of various cancer datasets makes this issue still challenging. In this paper, a novel Bio-inspired Multi-objective algorithm is proposed for gene selection in microarray data classification specifically in the binary domain of feature selection. The presented method extends the traditional Bat Algorithm with refined formulations, effective multi-objective operators, and novel local search strategies employing social learning concepts in designing random walks. A hybrid model using the Fisher criterion is then applied to three widely-used microarray cancer datasets to explore significant biomarkers which reveal the effectiveness of the proposed method for genomic analysis. Experimental results unveil new combinations of informative biomarkers have association with other studies.
The hospital outpatient non-attendance imposes huge financial burden on hospitals every year. The nonattendance issue roots in multiple diverse reasons which makes the problem space particularly complicated and undiscovered. The aim of this research is to build an advanced predictive model for non-attendance considering whole spectrum of factors and their complexities from big hospital data. We proposed a novel non-attendance prediction model based on deep neural networks. The proposed method is based on sparse stacked denoising autoencoders (SSDAEs). Different with exiting deep learning applications in hospital data which have separated data reconstruction and prediction phases, our model integrated both phases aiming to have higher performance than dividedclassification model in predicting tasks from EPR. The proposed method is compared with some well-known machine learning classifiers and representative research works for non-attendance prediction. The evaluation results reveal that the proposed deep approach drastically outperforms other methods in practice.
The hospital outpatient non-attendance imposes a substantial financial burden on hospitals and roots in multiple diverse reasons. This research aims to build an advanced predictive model for predicting non-attendance regarding the whole spectrum of probable contributing factors to non-attendance that could be collated from heterogeneous sources including electronic patients records and external non-hospital data. We proposed a new non-attendance prediction model based on deep neural networks and machine learning models. The proposed approach works upon sparse stacked denoising autoencoders (SDAEs) to learn the underlying manifold of data and thereby compacting information and providing a better representation that can be utilised afterwards by other learning models as well. The proposed approach is evaluated over real hospital data and compared with several well-known and scalable machine learning models. The evaluation results reveal the proposed approach with softmax layer and logistic regression outperforms other methods in practice.
In this paper, a general approach for international vehicle license plate localization and recognition is proposed. A hybrid solution is presented with combining basic machine vision techniques and neural networks. The proposed model consists of three main parts, including localization, segmentation and recognition. In the license plate localization, after some essential preprocessing and finding edges, the 8-connectivity of image background eliminates which helps more appropriately separating of main image objects from the cluttered backgrounds. Then, it is tried to find connected objects with 8-connectivity of the differentiated binary image. The binarization of license plate is based on local binarizing. The proposed recognizing system utilizes the Hough transform, basic morphological operators and Skeletonizing to provide an appropriate input for artificial neural networks. Segment by segment, the input streams into an intelligent error control unit (IECU) which itself is an already trained multi-layer perceptron (MLP) neural network. IECU investigates empty or non-character-inside segments. In case of no error, each segment streams into two already trained MLPs. Each of them singly recognizes either the alphabets or numbers. We show that this approach achieves accuracy over 91% on localizing vehicle license plate. The image database includes images of various vehicles with different background and slop under varying illumination conditions. The character recognition system correctly recognizes alphabets with probability over 97% and over 94% in case of numbers.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.