Browsing by Subject "Support vector machines (SVM)"
Now showing 1 - 3 of 3
Results Per Page
Sort Options
Item Face authentication with pose adjustment using support vector machines with a Hausdorff-based kernel(Texas Tech University, 2007-12) Wagner, Gregory M.; Sinzinger, Eric D.; Youn, Eunseog; Mengel, Susan A.; Hoo, Karlene A.Face authentication is a biometric classification method that verifies the identity of a user based an image of their face. Accuracy of the authentication is reduced when the pose of the training face images is different than the testing image. This dissertation describes two methodologies which can increase the face authentication accuracy, if the training and testing images poses are different. The first method uses cascading trilinear tensors which adjust the pose of 2D images in a 3D space. By being able to morph the images in a 3D space, the training and testing images can be normalized to have the same pose. Using support vector machines (SVM) as the classifier, the second method uses a Hausdorff-based kernel embedded in the SVM decision function. The Hausdorff-based kernel has been shown to improve accuracy in object recognition. Using these two methods, the face authentication accuracy is improved over methods which use classic SVM kernels or do not use pose adjustment.Item Feature evaluation of the support vector machine for micro-RNA target prediction in Arabidopsis thaliana based on antisense transcription and small RNA abundance(2007-12) Gontcharova, Viktoria; Youn, Eunseog; Rock, Chris; Watson, RichardMicro RNAs (miRNAs) are small non coding RNA that contribute to post transcriptional regulation. They are 21-23 nucleotide long sequences that effect development by binding by Watson- Crick pairing to a target gene and antagonizing various pathways of expression. This thesis explores the miRNA binding within the Arabidopsis thaliana genome as it relates to antisense transcription of target genes. Presented is a prediction mechanism that is based on two related features antisense transcription and small RNA abundance, hypothesized to be markers of the miRNA binding site in the target gene. A newly discovered phenomenon in the antisense strand of the target genes was implemented as a novel feature for target gene prediction. This feature, along with small RNAs and a commonly used indicator of binding sites, were used in a Support Vector Machine to build a prediction model. The three features were incorporated and analyzed using the output of the Support Vector Machine. Comparison was made between predicted and validated classifications to evaluate the importance of the features. Based on the accuracy, specificity, sensitivity and precision of the SVM results, the newly discovered feature may be able to identify new miRNA target sires in Arabidopsis and other species with deep genomic resources.Item Implementing artificial neural networks and support vector machines in stuck pipe prediction(2012-08) Albaiyat, Islam; Heinze, Lloyd R.; Siddiqui, Shameem; Youn, Eunseog; Beruvides, Mario G.Stuck pipe has been recognized as one of the most challenging and costly problems in the oil and gas industry. However, this problem can be treated proactively through predicting it before it occurs. The purpose of this study is to implement the two most powerful machine learning methods, Artificial Neural Networks (ANNs) and Support Vector Machines (SVMs), to predict stuck pipe occurrences. Two developed models for ANNs and SVMs with different scenarios were implemented for prediction purposes. The models were designed and constructed by the MATLAB language. The MATLAB built-in functions of ANNs and SVMs, and the MATLAB interface from the library of support vector machines were applied to compare the results. Furthermore, two databases that include mud properties, directional characteristics, and drilling parameters have been assembled for the training and testing processes. The study involved classifying stuck pipe incidents into two groups - stuck and non-stuck - and also into three groups: differentially stuck, mechanically stuck, and non-stuck. This research also has gone through optimization process which is vital in machine learning techniques to construct the most practical models. It showed that both ANNs and SVMs are able to predict stuck pipe occurrences with a reasonable accuracy which is over 83%. It has been shown in this study that the competitive SVM technique is able to generate promising and reasonable results of stuck pipe prediction. Besides, it can be found that SVMs are more convenient than ANNs since they need one or two parameters at most to be optimized. The constructed models generally apply very well in the areas for which they are built but may not work for other areas. However, they are important especially when it comes to probability measures. Thus, they can be utilized with real-time data and would represent the results on a log viewer.