No abstract

Abstract. Outlier detection has recently become an important problem in many industrial and financial applications. In this paper, a novel unsupervised algorithm for outlier detection with a solid statistical foundation is proposed. First we modify a nonparametric density estimate with a variable kernel to yield a robust local density estimation. Outliers are then detected by comparing the local density of each point to the local density of its neighbors. Our experiments performed on several simulated data sets have demonstrated that the proposed approach can outperform two widely used outlier detection algorithms (LOF and LOCI).

Purpose: The authors present an efficient method for generating anthropomorphic software breast phantoms with high spatial resolution. Employing the same region growing principles as in their previous algorithm for breast anatomy simulation, the present method has been optimized for computational complexity to allow for fast generation of the large number of phantoms required in virtual clinical trials of breast imaging. Methods: The new breast anatomy simulation method performs a direct calculation of the Cooper's ligaments (i.e., the borders between simulated adipose compartments). The calculation corresponds to quadratic decision boundaries of a maximum a posteriori classifier. The method is multiscale due to the use of octree-based recursive partitioning of the phantom volume. The method also provides user-control of the thickness of the simulated Cooper's ligaments and skin. Results: Using the proposed method, the authors have generated phantoms with voxel size in the range of (25-1000 lm)

Reliable diagnosis of depressive disorder is essential for both optimal treatment and prevention of fatal outcomes. This study aimed to elucidate the effectiveness of two non-linear measures, Higuchi's Fractal Dimension (HFD) and Sample Entropy (SampEn), in detecting depressive disorders when applied on EEG. HFD and SampEn of EEG signals were used as features for seven machine learning algorithms including Multilayer Perceptron, Logistic Regression, Support Vector Machines with the linear and polynomial kernel, Decision Tree, Random Forest, and Naı ¨ve Bayes classifier, discriminating EEG between healthy control subjects and patients diagnosed with depression. This study confirmed earlier observations that both non-linear measures can discriminate EEG signals of patients from healthy control subjects. The results suggest that good classification is possible even with a small number of principal components. Average accuracy among classifiers ranged from 90.24 to 97.56%. Among the two measures, SampEn had better performance. Using HFD and SampEn and a variety of machine learning techniques we can accurately discriminate patients diagnosed with depression vs controls which can serve as a highly sensitive, clinically relevant marker for the diagnosis of depressive disorders.

Outlier detection has recently become an important problem in many data mining applications. In this paper, a novel unsupervised algorithm for outlier detection is proposed. First we apply a provably globally optimal Expectation Maximization (EM) algorithm to fit a Gaussian Mixture Model (GMM) to a given data set. In our approach, a Gaussian is centered at each data point, and hence, the estimated mixture proportions can be interpreted as probabilities of being a cluster center for all data points. The outlier factor at each data point is then defined as a weighted sum of the mixture proportions with weights representing the similarities to other data points. The proposed outlier factor is thus based on global properties of the data set. This is in contrast to most existing approaches to outlier detection, which are strictly local. Our experiments performed on several simulated and real life data sets demonstrate superior performance of the proposed approach. Moreover, we also demonstrate the ability to detect unusual shapes.

The calcar femorale (CF), a plate of dense bone internal to the lesser trochanter, is visible on computed tomographic images of the 6 million‐year‐old femoral fragment BAR 1003′00 (from the taxon Orrorin tugenensis), among the oldest specimens relevant to reconstructing the evolution of human bipedal locomotion. A strongly expressed CF has been used previously as an indicator of bipedality. If true, then there should be a quantifiable difference in the CF among hominoids. Absolute and normalized CF lengths were measured from computed tomographic images at five anatomical locations along the proximal portion of BAR 1003′00 in addition to samples of nine H. sapiens and ten P. troglodytes femora. The span of the CF superiorly to inferiorly within the proximal femur was measured by counting the number of cross‐sections on which the CF occurred. A Bayesian approach was used to classify the BAR 1003′00 sample based on normalized lengths. The P. troglodytes femora were more variable both in the occurrence of the trait and, where present, its span in the proximal femur. The H. sapiens sample exhibited CF lengths that were consistently larger at each location than the P. troglodytes in absolute terms, but the normalized lengths overlap substantially. The Bayesian posterior probability test classifies the CF of BAR 1003′00 with H. sapiens. The BAR 1003′00’s calcar femorale has a strong anatomical similarity to the H. sapiens sample, supporting the conclusion that O. tugenensis is an early bipedal hominin. Anat Rec, 301:1834–1839, 2018. © 2018 Wiley Periodicals, Inc.

No abstract

We perform multi-class classification of laser-induced breakdown spectroscopy data of four commercial samples of proteins diluted in phosphate-buffered saline solution at different concentrations: bovine serum albumin, osteopontin, leptin, and insulin-like growth factor II. We achieve this by using principal component analysis as a method for dimensionality reduction. In addition, we apply several different classification algorithms (K-nearest neighbor, classification and regression trees, neural networks, support vector machines, adaptive local hyperplane, and linear discriminant classifiers) to perform multi-class classification. We achieve classification accuracies above 98% by using the linear classifier with 21-31 principal components. We obtain the best detection performance for neural networks, support vector machines, and adaptive local hyperplanes for a range of the number of principal components with no significant differences in performance except for that of the linear classifier. With the optimal number of principal components, a simplistic K-nearest classifier still provided acceptable results. Our proposed approach demonstrates that highly accurate automatic classification of complex protein samples from laser-induced breakdown spectroscopy data can be successfully achieved using principal component analysis with a sufficiently large number of extracted features, followed by a wrapper technique to determine the optimal number of principal components.