Optimized Support Vector Machine Classifier Research Articles

HHO optimized support vector machine classifier for traditional Chinese medicine syndrome differentiation of diabetic retinopathy.

To develop a classifier for traditional Chinese medicine (TCM) syndrome differentiation of diabetic retinopathy (DR), using optimized machine learning algorithms, which can provide the basis for TCM objective and intelligent syndrome differentiation. Collated data on real-world DR cases were collected. A variety of machine learning methods were used to construct TCM syndrome classification model, and the best performance was selected as the basic model. Genetic Algorithm (GA) was used for feature selection to obtain the optimal feature combination. Harris Hawk Optimization (HHO) was used for parameter optimization, and a classification model based on feature selection and parameter optimization was constructed. The performance of the model was compared with other optimization algorithms. The models were evaluated with accuracy, precision, recall, and F1 score as indicators. Data on 970 cases that met screening requirements were collected. Support Vector Machine (SVM) was the best basic classification model. The accuracy rate of the model was 82.05%, the precision rate was 82.34%, the recall rate was 81.81%, and the F1 value was 81.76%. After GA screening, the optimal feature combination contained 37 feature values, which was consistent with TCM clinical practice. The model based on optimal combination and SVM (GA_SVM) had an accuracy improvement of 1.92% compared to the basic classifier. SVM model based on HHO and GA optimization (HHO_GA_SVM) had the best performance and convergence speed compared with other optimization algorithms. Compared with the basic classification model, the accuracy was improved by 3.51%. HHO and GA optimization can improve the model performance of SVM in TCM syndrome differentiation of DR. It provides a new method and research idea for TCM intelligent assisted syndrome differentiation.

Text Mining - A Comparative Review of Twitter Sentiments Analysis

Background: Text mining derives information and patterns from textual data. Online social media platforms, which have recently acquired great interest, generate vast text data about human behaviors based on their interactions. This data is generally ambiguous and unstructured. The data includes typing errors and errors in grammar that cause lexical, syntactic, and semantic uncertainties. This results in incorrect pattern detection and analysis. Researchers are employing various text mining techniques that can aid in Topic Modeling, the detection of Trending Topics, the identification of Hate Speeches, and the growth of communities in online social media networks. Objective: This review paper compares the performance of ten machine learning classification techniques on a Twitter data set for analyzing users' sentiments on posts related to airline usage. Methods: Review and comparative analysis of Gaussian Naive Bayes, Random Forest, Multinomial Naive Bayes, Multinomial Naive Bayes with Bagging, Adaptive Boosting (AdaBoost), Optimized AdaBoost, Support Vector Machine (SVM), Optimized SVM, Logistic Regression, and Long-Short Term Memory (LSTM) for sentiment analysis. Results: The results of the experimental study showed that the Optimized SVM performed better than the other classifiers, with a training accuracy of 99.73% and testing accuracy of 89.74% compared to other models. Conclusion: Optimized SVM uses the RBF kernel function and nonlinear hyperplanes to split the dataset into classes, correctly classifying the dataset into distinct polarity. This, together with Feature Engineering utilizing Forward Trigrams and Weighted TF-IDF, has improved Optimized SVM classifier performance regarding train and test accuracy. Therefore, the train and test accuracy of Optimized SVM are 99.73% and 89.74% respectively. When compared to Random Forest, a marginal of 0.09% and 1.73% performance enhancement is observed in terms of train and test accuracy and 1.29% (train accuracy) and 3.63% (test accuracy) of improved performance when compared with LSTM. Likewise, Optimized SVM, gave more than 10% of enhanced performance in terms of train accuracy when compared with Gaussian Naïve Bayes, Multinomial Naïve Bayes, Multinomial Naïve Bayes with Bagging, Logistic Regression and a similar enhancement is observed with AdaBoost and Optimized AdaBoost which are ensemble models during the experimental process. Optimized SVM also has outperformed all the classification models in terms of AUC-ROC train and test scores.

Multi-Classification of Electroencephalogram Epileptic Seizures Based on Robust Hybrid Feature Extraction Technique and Optimized Support Vector Machine Classifier

Multi-Classification of Electroencephalogram Epileptic Seizures Based on Robust Hybrid Feature Extraction Technique and Optimized Support Vector Machine Classifier

Automated Brain Tumour Detection and Classification using Deep Features and Bayesian Optimised Classifiers.

Brain tumour detection and classification require trained radiologists for efficient diagnosis. The proposed work aims to build a Computer Aided Diagnosis (CAD) tool to automate brain tumour detection using Machine Learning (ML) and Deep Learning (DL) techniques. Magnetic Resonance Image (MRI) collected from the publicly available Kaggle dataset is used for brain tumour detection and classification. Deep features extracted from the global pooling layer of Pretrained Resnet18 network are classified using 3 different ML Classifiers, such as Support vector Machine (SVM), K-Nearest Neighbour (KNN), and Decision Tree (DT). The above classifiers are further hyperparameter optimised using Bayesian Algorithm (BA) to enhance the performance. Fusion of features extracted from shallow and deep layers of the pretrained Resnet18 network followed by BA-optimised ML classifiers is further used to enhance the detection and classification performance. The confusion matrix derived from the classifier model is used to evaluate the system's performance. Evaluation metrics, such as accuracy, sensitivity, specificity, precision, F1 score, Balance Classification Rate (BCR), Mathews Correlation Coefficient (MCC) and Kappa Coefficient (Kp), are calculated. Maximum accuracy, sensitivity, specificity, precision, F1 score, BCR, MCC, and Kp of 99.11 %, 98.99 %, 99.22 %, 99.09 %, 99.09 %, 99.10 %, 98.21 %, 98.21 %, respectively, were obtained for detection using fusion of shallow and deep features of Resnet18 pretrained network classified by BA optimized SVM classifier. Feature fusion performs better for classification task with accuracy, sensitivity, specificity, precision, F1 score, BCR, MCC and Kp of 97.31 %, 97.30 %, 98.65 %, 97.37 %, 97.34 %, 97.97%, 95.99 %, 93.95 %, respectively. The proposed brain tumour detection and classification framework using deep feature extraction from Resnet 18 pretrained network in conjunction with feature fusion and optimised ML classifiers can improve the system performance. Henceforth, the proposed work can be used as an assistive tool to aid the radiologist in automated brain tumour analysis and treatment.

EEG asymmetry detection in patients with severe acquired brain injuries via machine learning methods

Lateral brain symmetry indexes, detected by electroencephalography (EEG), are markers of rehabilitative recovery widely used in patients with severe acquired brain injury (sABI). In this study, Machine Learning algorithms were cross-validated to detect consistent asymmetries, starting from a completely automated features extraction pipeline in the EEG recordings of 54 patients with sABI, classified by two expert neurophysiologists. Raw data were filtered and segmented in two-seconds non-overlapping epochs. Low data quality in frontal electrodes caused up to 40% of epochs rejection, whilst central and posterior electrodes contributed with the greatest number of artefacts-free epochs. Out of more than 3000 extracted features, ∼300 significantly differentiated symmetric and asymmetric EEG recordings, most of them extracted from pairs and lines of electrodes. The best performing solution (nested-cross-validated and optimized Support Vector Machine classifier) detected asymmetry with a test accuracy of 85% (sensitivity 92%, specificity 80%). The application of the proposed approach to our sample size supports the generalizability of our model and its translation to clinical practice. The algorithm, heading to automatic EEG analysis, has the potential to speed up analysis of long recordings and, to improve quality of care in settings lacking skilled staff.

An optimized technique for copy–move forgery localization using statistical features

Copy–Move Forgery Detection (CMFD) helps to detect copied and pasted areas in one image. It plays a crucial role in legal evidence, forensic investigation, defence, and many more places. In the proposed CMFD method, a two-step identification of forgery is presented. In step one, the suspected image will be classified into either one of two classes that are forged or authentic. Step two is carried out​ only if the suspected is classified as forged, then forged location will be identified using the block-matching procedure. Initially, the suspected image is decomposed into different orientations using Steerable Pyramid Transform (SPT); Grey Level Co-occurrence Matrix (GLCM) features are extracted from each orientation. These features are used to train Optimized Support Vector Machine (OSVM) as well as to classify. If the suspected image is categorized into forged, then the suspected grey image is converted into overlapping blocks, and from each block, GLCM features are extracted. The proper similarity threshold value and distance threshold value can locate the forged region using GLCM block features. The performance of the proposed method is tested using standard datasets CoMoFoD and CASIA Datasets. The proposed CMFD approach results are consistent, even the forged image suffered from attacks like JPEG compression, scaling, and rotation. The OSVM classifier is showing superiority over the Optimized Naive Bayes Classifier (ONBC), Extreme Learning Machine (ELM) and Support Vector Machine (SVM).

Hybrid Feature-Based Disease Detection in Plant Leaf Using Convolutional Neural Network, Bayesian Optimized SVM, and Random Forest Classifier

Plant diseases are unfavourable factors that cause a significant decrease in the quality and quantity of crops. Experienced biologists or farmers often observe plants with the naked eye for disease, but this method is often imprecise and can take a long time. In this study, we use artificial intelligence and computer vision techniques to achieve the goal of designing and developing an intelligent classification mechanism for leaf diseases. This paper follows two methodologies and their simulation outcomes are compared for performance evaluation. In the first part, data augmentation is performed on the PlantVillage data set images (for apple, corn, potato, tomato, and rice plants), and their deep features are extracted using convolutional neural network (CNN). These features are classified by a Bayesian optimized support vector machine classifier and the results attained in terms of precision, sensitivity, f-score, and accuracy. The above-said methodologies will enable farmers all over the world to take early action to prevent their crops from becoming irreversibly damaged, thereby saving the world and themselves from a potential economic crisis. The second part of the methodology starts with the preprocessing of data set images, and their texture and color features are extracted by histogram of oriented gradient (HoG), GLCM, and color moments. Here, the three types of features, that is, color, texture, and deep features, are combined to form hybrid features. The binary particle swarm optimization is applied for the selection of these hybrid features followed by the classification with random forest classifier to get the simulation results. Binary particle swarm optimization plays a crucial role in hybrid feature selection; the purpose of this Algorithm is to obtain the suitable output with the least features. The comparative analysis of both techniques is presented with the use of the above-mentioned evaluation parameters.

Distinguishing patients with temporal lobe epilepsy from normal controls with the directed graph measures of resting-state fMRI

PurposeChanges in the brain networks of patients with temporal lobe epilepsy (TLE) have been extensively explored, but the biological mechanisms underlying these alterations remain unclear. Here, we aim to identify changes in brain networks in patients with TLE and provide an accurate algorithm for distinguishing these patients from normal controls (NC) with graph-theoretical approach and advanced machine learning methods. MethodsDirected network construction was applied to resting-state functional magnetic resonance imaging (rs-fMRI) data from 55 subjects (23 TLE patients and 32 NC), and 13 directed graph measures were calculated. Two-sample t-test selected features were used as inputs to a support vector machine (SVM). The leave-one-out cross-validation method was used in measuring classification performance. ResultsAn accuracy of 94.55% (sensitivity = 91.30%, f1-score = 93.33%, Cohen's kappa = 0.9345) was achieved for the classification of TLE patients and NC with optimal features and SVM classifier. According to the results of the two-sample t-test results, TLE disease impacted several areas of the brain, including the temporal, parietal, occipital, posterior cingulate, angular gyrus, superior frontal gyrus, and cerebellum regions in degree centrality, flow coefficient and node efficiency. There was a significant correlation between performance IQ and the flow coefficient of the left posterior cerebellum lobe in TLE group. ConclusionThe study confirmed the validity of Granger causality analysis in constructing directed brain networks. The proposed machine learning approach based on directed graph measures may serve as a biomarker for the diagnosis of TLE to assist in the early diagnosis of TLE patients and intervention in treatment plans.

Ensembles of Deep Learning Framework for Stomach Abnormalities Classification

Abnormalities of the gastrointestinal tract are widespread worldwide today. Generally, an effective way to diagnose these life-threatening diseases is based on endoscopy, which comprises a vast number of images. However, the main challenge in this area is that the process is time-consuming and fatiguing for a gastroenterologist to examine every image in the set. Thus, this led to the rise of studies on designing AI-based systems to assist physicians in the diagnosis. In several medical imaging tasks, deep learning methods, especially convolutional neural networks (CNNs), have contributed to the state-of-the-art outcomes, where the complicated nonlinear relation between target classes and data can be learned and not limit to hand-crafted features. On the other hand, hyperparameters are commonly set manually, which may take a long time and leave the risk of non-optimal hyperparameters for classification. An effective tool for tuning optimal hyperparameters of deep CNN is Bayesian optimization. However, due to the complexity of the CNN, the network can be regarded as a black-box model where the information stored within it is hard to interpret. Hence, Explainable Artificial Intelligence (XAI) techniques are applied to overcome this issue by interpreting the decisions of the CNNs in such wise the physicians can trust. To play an essential role in real-time medical diagnosis, CNN-based models need to be accurate and interpretable, while the uncertainty must be handled. Therefore, a novel method comprising of three phases is proposed to classify these life-threatening diseases. At first, hyperparameter tuning is performed using Bayesian optimization for two state-of-the-art deep CNNs, and then Darknet53 and InceptionV3 features are extracted from these fine-tunned models. Secondly, XAI techniques are used to interpret which part of the images CNN takes for feature extraction. At last, the features are fused, and uncertainties are handled by selecting entropy-based features. The experimental results show that the proposed method outperforms existing methods by achieving an accuracy of 97% based on a Bayesian optimized Support Vector Machine classifier.

Extracting Function-Driven Tracing Characteristics for Optimized SVM Classification

Due to its openness and simplicity, Modbus TCP has wide applications to facilitate the actual management and control in industrial wireless fields. However, its potential security vulnerabilities can also create lots of complicated information security challenges, which are increasingly threatening the availability of industrial real-time traffic delivery. Although anomaly detection has been recognized as a workable security measure to identify attacks, the critical step to successfully extract data characteristics is an extremely difficult task. In this paper, we focus on the continuous control mode in industrial processes and propose a control tracing feature algorithm to extract the function-driven tracing characteristics from Modbus TCP data traffic. Furthermore, this algorithm can flexibly integrate the time factor with critical functional operations and adequately describe the dynamic control change of technological processes. To closely cooperate with this algorithm, one optimized SVM (support vector machine) classifier is introduced as the practicable decision engine. By designing one applicable attack mode, we develop an in-depth and meticulous analysis on the decision accuracy, and all experimental results clearly explain that the extracted features can strongly reflect the changing pattern of continuous functional operations, and the proposed algorithm can effectively cooperate with the optimized SVM classifier to distinguish abnormal Modbus TCP data traffic.

Attentional load classification in multiple object tracking task using optimized support vector machine classifier: a step towards cognitive brain–computer interface

Cognitive brain–computer interface (cBCI) is an emerging area with applications in neurorehabilitation and performance monitoring. cBCI works on the cognitive brain signal that does not require a person to pay much effort unlike the motor brain-computer interface (BCI) however existing cBCI systems currently offer lower accuracy than the motor BCI. Since attention is one of the cognitive signals that can be used to realise the cBCI, this work uses the multiple object tracking (MOT) task to acquire the desired electroencephalograph (EEG) signal from healthy subjects. The main objective of the paper is to explore the preliminary applications of support vector machine (SVM) classifier to classify the attentional load in multiple object tracking task. Results show that the attentional load can be classified using SVM with sensitivity, specificity, and accuracy of 94.03%, 92.50%, and 93.28%, respectively using the spectral entropy EEG feature. The classification performance promises the potential application of the current approach in the cognitive brain-computer interface for neurorehabilitation.

Prediction of genetically-evaluated tumour responses to chemotherapy from breast MRI using machine learning with model selection

Genetic tests can provide prognostic information in breast cancer for both diagnosis and treatment planning. However, the cost of a genetic test is still high. In this study, we developed a radiogenomics method to predict genetically-evaluated responses to chemotherapy for breast cancer using our machine-learning technology coupled with model selection. Our proposed method consists of feature extraction, model selection, and prediction by the selected model. In the feature extraction, 318 morphological and texture features were extracted from a tumour region. In the model selection module, there are two major components: (1) selection of imaging biomarkers based on our original sequential forward floating selection (SFFS) feature selection and (2) building of a support vector machine (SVM) classifier including kernel function selection and hyperparameter optimization. The optimized feature set, i.e. imaging biomarkers, coupled with an SVM classifier were chosen by maximizing the area under curve (AUC) of corresponding receiver-operating-characteristic curve (ROC). After the model selection, the optimized SVM classifier operated on the selected imaging biomarkers for prediction. We applied our proposed method to 118 breast MRI studies from 118 patients for predicting genetically-evaluated responses to chemotherapy for breast cancer that evaluated by the genetic test of IRSN-23. We achieved an AUC value of 0.96 using the optimized SVM classifier model coupled with 24 selected imaging biomarkers in predicting the results of IRSN-23 in a five-fold cross-validation procedure.

Colon cancer prediction on histological images using deep learning features and Bayesian optimized SVM

Colon cancer is one of the highest cancer diagnosis mortality rates worldwide. However, relying on the expertise of pathologists is a demanding and time-consuming process for histopathological analysis. The automated diagnosis of colon cancer from biopsy examination played an important role for patients and prognosis. As conventional handcrafted feature extraction requires specialized experience to select realistic features, deep learning processes have been chosen as abstract high-level features may be extracted automatically. This paper presents the colon cancer detection system using transfer learning architectures to automatically extract high-level features from colon biopsy images for automated diagnosis of patients and prognosis. In this study, the image features are extracted from a pre-trained convolutional neural network (CNN) and used to train the Bayesian optimized Support Vector Machine classifier. Moreover, Alexnet, VGG-16, and Inception-V3 pre-trained neural networks were used to analyze the best network for colon cancer detection. Furthermore, the proposed framework is evaluated using four datasets: two are collected from Indian hospitals (with different magnifications 4X, 10X, 20X, and 40X) and the other two are public colon image datasets. Compared with the existing classifiers and methods using public datasets, the test results evaluated the Inception-V3 network with the accuracy range from 96.5% - 99% as best suited for the proposed framework.

A Method for Constructing Automatic Rolling Bearing Fault Identification Model Based on Refined Composite Multi-Scale Dispersion Entropy

In this paper, one of most widely utilized rolling bearings in rotating machinery is selected as the research object. Automatic rolling bearing fault identification model including support vector machine (SVM) training module, fault classification knowledge base module, and fault automatic identification module is proposed. A generalized method for automatic identification of rolling bearing faults based on refined composite multi-scale dispersion entropy (RCMDE) is developed. First, in order to solve the problem of setting the value range of the decomposition level K based on empirical knowledge for variational modal decomposition (VMD), a maximum kurtosis value method is proposed to determine the preset value range of K in whale optimization algorithm. Then, an improved VMD method is used to adaptively decompose the signal into a series of intrinsic mode function components. Next, the correlation coefficient method is employed to screen effective feature components of bearings in different health states for reconstruction. Through theoretical analysis, the calculated RCMDE value of reconstructed signal is screened and input as a feature value into the optimized SVM classifier for fault pattern recognition. The input of rolling bearing vibration data without preprocessing and the output of the fault identification which don't rely on empirical knowledge of external experts is realized. Experimental and engineering case data of rolling bearings under different equipment and operating environments are tested and validated. The results indicate that the model proposed in this paper shows good fault identification, demonstrates good generalization performance, and has beneficial industrial application prospect.

Genetic algorithm–optimized support vector machine for real-time activity recognition in health smart home

Health smart home, as a typical application of Internet of things, provides a new solution for remote medical treatment. It can effectively relieve pressure from shortage of medical resources caused by aging population and help elderly people live at home more independently and safely. Activity recognition is the core of health smart home. This technology aims to recognize the activity patterns of users from a series of observations on the user’ actions and the environmental conditions, so as to avoid distress situations as much as possible. However, most of the existing researches focus on offline activity recognition, but not good at online real-time activity recognition. Besides, the feature representation techniques used for offline activity recognition are generally not suitable for online scenarios. In this article, the authors propose a real-time online activity recognition approach based on the genetic algorithm–optimized support vector machine classifier. In order to support online real-time activity recognition, a new sliding window-based feature representation technique enhanced by mutual information between sensors is devised. In addition, the genetic algorithm is used to automatically select optimal hyperparameters for the support vector machine model, thereby reducing the recognition inaccuracy caused by manual tuning of hyperparameters. Finally, a series of comprehensive experiments are conducted on freely available data sets to validate the effectiveness of the proposed approach.

Hybrid Learning of Hand-Crafted and Deep-Activated Features Using Particle Swarm Optimization and Optimized Support Vector Machine for Tuberculosis Screening

Tuberculosis (TB) is a leading infectious killer, especially for people with Human Immunodeficiency Virus (HIV) and Acquired Immunodeficiency Syndrome (AIDS). Early diagnosis of TB is crucial for disease treatment and control. Radiology is a fundamental diagnostic tool used to screen or triage TB. Automated chest x-rays analysis can facilitate and expedite TB screening with fast and accurate reports of radiological findings and can rapidly screen large populations and alleviate a shortage of skilled experts in remote areas. We describe a hybrid feature-learning algorithm for automatic screening of TB in chest x-rays: it first segmented the lung regions using the DeepLabv3+ model. Then, six sets of hand-crafted features from statistical textures, local binary pattern, GIST, histogram of oriented gradients (HOG), pyramid histogram of oriented gradients and bags of visual words (BoVW), and nine sets of deep-activated features from AlexNet, GoogLeNet, InceptionV3, XceptionNet, ResNet-50, SqueezeNet, ShuffleNet, MobileNet, and DenseNet, were extracted. The dominant features of each feature set were selected using particle swarm optimization, and then separately input to an optimized support vector machine classifier to label ‘normal’ and ‘TB’ x-rays. GIST, HOG, BoVW from hand-crafted features, and MobileNet and DenseNet from deep-activated features performed better than the others. Finally, we combined these five best-performing feature sets to build a hybrid-learning algorithm. Using the Montgomery County (MC) and Shenzen datasets, we found that the hybrid features of GIST, HOG, BoVW, MobileNet and DenseNet, performed best, achieving an accuracy of 92.5% for the MC dataset and 95.5% for the Shenzen dataset.

Investigating the gene expression profiles of cells in seven embryonic stages with machine learning algorithms

The development of embryonic cells involves several continuous stages, and some genes are related to embryogenesis. To date, few studies have systematically investigated changes in gene expression profiles during mammalian embryogenesis. In this study, a computational analysis using machine learning algorithms was performed on the gene expression profiles of mouse embryonic cells at seven stages. First, the profiles were analyzed through a powerful Monte Carlo feature selection method for the generation of a feature list. Second, increment feature selection was applied on the list by incorporating two classification algorithms: support vector machine (SVM) and repeated incremental pruning to produce error reduction (RIPPER). Through SVM, we extracted several latent gene biomarkers, indicating the stages of embryonic cells, and constructed an optimal SVM classifier that produced a nearly perfect classification of embryonic cells. Furthermore, some interesting rules were accessed by the RIPPER algorithm, suggesting different expression patterns for different stages.

A new intelligent fault diagnosis method for bearing in different speeds based on the FDAF-score algorithm, binary particle swarm optimization, and support vector machine

In this paper, a new hybrid intelligent technique is presented based on the improvement in the feature selection method for multi-fault classification. The bearing conditions used in this study include healthy condition, defective inner ring, defective outer ring, and the faulty rolling element at different rotating motor speeds. To form the feature matrix, at first, the vibration signals are decomposed using empirical mode decomposition and wavelet packet decomposition. Then, the time and frequency domain features are extracted from the raw signals and the components are obtained from the signal decomposition. The high-dimensional feature matrix leads to increasing the computational complexity and reducing the efficiency in the classification accuracy of faults. Therefore, in the first stage of the feature selection process, the redundant and unnecessary features are eliminated by the FDAF-score feature selection method and the preselected feature set is formed. The FDAF-score technique is a combination of both F-score and Fisher discriminate analysis (FDA) algorithms. Since there may exist the features that are not susceptible to the presence of faults, the binary particle swarm optimization (BPSO) algorithm and the support vector machine (SVM) are used to select the optimal features from the preselected features. The BPSO algorithm is used to determine the optimal feature set and SVM classifier parameters so that the predictive error of the bearing conditions and the number of selected features are minimized. The results obtained in this paper demonstrate that the selected features are able to differentiate the different bearing conditions at various speeds. Comparing the results of this article with other fault detection methods indicates the ability of the proposed method.

Detection of fraud risks in retailing sector using MLP and SVM techniques

In today's business conditions, where business activities are spreading over a wide geographical area, fraud auditing processes have crucial importance especially for the retailing sector which has a high branch network. In the retailing sector, especially purchasing processes are subject to high fraud risks. This paper shows that it is possible to detect fraudulent processes by applying data mining techniques on operational data related to purchasing activities. Within this scope, in order to detect the fraudulent purchasing operations, support vector machine (SVM) models with different kernels and artificial neural networks methods have been used and successful results have been achieved. The results of the two methods have been examined comparatively and it shows that optimized SVM classifier outperforms others. Besides, in this study, it is presumed that the detected fraud data can be proactively used in the struggle against fraud with fraud-governance risk and compliance software by converting it into scenario analysis.

Primary Tumor Site Specificity is Preserved in Patient-Derived Tumor Xenograft Models.

Patient-derived tumor xenograft (PDX) mouse models are widely used for drug screening. The underlying assumption is that PDX tissue is very similar with the original patient tissue, and it has the same response to the drug treatment. To investigate whether the primary tumor site information is well preserved in PDX, we analyzed the gene expression profiles of PDX mouse models originated from different tissues, including breast, kidney, large intestine, lung, ovary, pancreas, skin, and soft tissues. The popular Monte Carlo feature selection method was employed to analyze the expression profile, yielding a feature list. From this list, incremental feature selection and support vector machine (SVM) were adopted to extract distinctively expressed genes in PDXs from different primary tumor sites and build an optimal SVM classifier. In addition, we also set up a group of quantitative rules to identify primary tumor sites. A total of 755 genes were extracted by the feature selection procedures, on which the SVM classifier can provide a high performance with MCC 0.986 on classifying primary tumor sites originated from different tissues. Furthermore, we obtained 16 classification rules, which gave a lower accuracy but clear classification procedures. Such results validated that the primary tumor site specificity was well preserved in PDX as the PDXs from different primary tumor sites were still very different and these PDX differences were similar with the differences observed in patients with tumor. For example, VIM and ABHD17C were highly expressed in the PDX from breast tissue and also highly expressed in breast cancer patients.