Search

The state of Wisconsin has one of the highest four year graduation rates in the nation, but deep disparities among student subgroups remain. To address this the state has created the Wisconsin Dropout Early Warning System (DEWS), a predictive model of student dropout risk for students in grades six through nine. The Wisconsin DEWS is in use statewide and currently provides predictions on the likelihood of graduation for over 225,000 students. DEWS represents a novel statistical learning based approach to the challenge of assessing the risk of non-graduation for students and provides highly accurate predictions for students in the middle grades without expanding beyond mandated administrative data collections. Similar dropout early warning systems are in place in many jurisdictions across the country. Prior research has shown that in many cases the indicators used by such systems do a poor job of balancing the trade off between correct classification of likely dropouts and false-alarm (Bowers et al., 2013). Building on this work, DEWS uses the receiver-operating characteristic (ROC) metric to identify the best possible set of statistical models for making predictions about individual students. This paper describes the DEWS approach and the software behind it, which leverages the open source statistical language R (R Core Team, 2013). As a result DEWS is a flexible series of software modules that can adapt to new data, new algorithms, and new outcome variables to not only predict dropout, but also impute key predictors as well. The design and implementation of each of these modules is described in detail as well as the open-source R package, EWStools, that serves as the core of DEWS (Knowles, 2014).

In this paper, we demonstrate that fixed-weight layers generated from random distribution or logistic functions can effectively extract significant features from input data, resulting in high accuracy on a variety of tasks, including Fisher's Iris, Wisconsin Breast Cancer, and MNIST datasets. We have observed that logistic functions yield high accuracy with less dispersion in results. We have also assessed the precision of our approach under conditions of minimizing the number of spikes generated in the network. It is practically useful for reducing energy consumption in spiking neural networks. Our findings reveal that the proposed method demonstrates the highest accuracy on Fisher's iris and MNIST datasets with decoding using logistic regression. Furthermore, they surpass the accuracy of the conventional (non-spiking) approach using only logistic regression in the case of Wisconsin Breast Cancer. We have also investigated the impact of non-stochastic spike generation on accuracy. [ABSTRACT FROM AUTHOR]

In recent decades, breast cancer has become one of the leading causes of mortality among women. This disease is not preventable because of its unknown causes; however, its early diagnosis increases patients' recovery chances. Machine learning (ML) can be utilized to improve treatment outcomes in healthcare operations while diminishing costs and time. In this research, we suggest two novel feature selection (FS) methods based upon an imperialist competitive algorithm (ICA) and a bat algorithm (BA) and their combination with ML algorithms. This study aims to enhance diagnostic models' efficiency and present a comprehensive analysis to help clinical physicians make more precise and reliable decisions. K-nearest neighbors (KNN), support vector machine (SVM), decision tree (DT), Naive Bayes, AdaBoost (AB), linear discriminant analysis (LDA), random forest (RF), logistic regression (LR), and artificial neural network (ANN) are some of the methods employed. Sensitivity, accuracy, precision, mean absolute error F-score, root mean square error, Kappa, and relative absolute error calculated the performance of the methods. This paper applied a distinctive integration of evaluation measures and ML algorithms using the wrapper feature selection based on ICA (WFSIC) and BA (WFSB) separately. We compared two proposed approaches for the performance of the classifiers. Also, we compared our best diagnostic model with previous works reported in the literature survey. Experimentations were performed on the Wisconsin diagnostic breast cancer (WDBC) dataset. Results reveal that the proposed framework that uses the BA with an accuracy of 99.12% surpasses the framework using the ICA and most previous works. Additionally, the RF classifier in the approach of FS based on BA emerges as the best model and outperforms others regarding its criteria. Besides, the results illustrate the role of our techniques in reducing the dataset dimensions up to 90% and increasing the performance of diagnostic models by over 99%. Moreover, the result demonstrates that there are more critical features than the optimum dataset obtained by proposed FS approaches that have been selected by most ML models, including the standard error of area, concavity, smoothness, perimeter, the worst of texture, compactness, radius, symmetry, smoothness, concavity, and the mean of concave points, fractal dimension, compactness, concavity that can remarkably affect the efficiency of breast cancer prediction. This study illustrates the role of our approaches in enhancing treatment outcomes in healthcare operations. [ABSTRACT FROM AUTHOR]

Recently, a system of spintronic vortex oscillators has been experimentally trained to classify vowel sounds. In this paper, we have carried out a combination of device-level and system-level simulations to train a system of spin Hall nano oscillators (SHNOs) of smaller size (25X lower in area compared to those vortex oscillators) for such data classification tasks. Magnetic moments precess in an uniform mode as opposed to the vortex mode in our oscillators. We have trained our system to classify inputs in various popular machine learning data sets like Fisher's Iris data set of flowers, Wisconsin Breast Cancer (WBC) data set, and MNIST data set of handwritten digits. We have employed a new technique for input dimensionality reduction here so that the clustering/target synchronization pattern changes based on the nature of the data in the different data sets. Our demonstration of learning in a system of such small SHNOs for a wide range of data sets is promising for scaling up the oscillator-based neuromorphic system for complex data classification tasks. [ABSTRACT FROM AUTHOR]

Breast cancer was the most diagnosed form of cancer in 2020. Early diagnosis of breast cancer results in a significant improvement in long-term survival rates. Current methods require consultation with experts, which is expensive and time-consuming and thus may not be accessible to all. This paper seeks to train and evaluate supervised machine learning models for the accurate and efficient detection of breast cancer. The Wisconsin Breast Cancer Database dataset describes 30 attributes of cell nuclei, including, but not limited to, their radius, texture, and concavity. It contains 569 instances, 212 of which are malignant tumors. The Random Forest algorithm outperforms other algorithms in classifying breast tumors as either malignant or benign and is thus selected as our primary model. It is trained on two different subsets of the dataset having 16 and 8 features, respectively, identified with the help of multiple feature selection methods. The Random Forest models are tested post hyperparameter tuning on a holdout set, and accuracies of 100% and 99.30% respectively. The models are also compared with four other machine learning classification algorithms: Support Vector Machine (SVM), Decision Tree, Multilayer Perceptron, and K-Nearest Neighbors. The results confirm that Random Forest is the superior method for breast cancer diagnosis. [ABSTRACT FROM AUTHOR]

Background: In classification and diagnostic testing, the receiver-operator characteristic (ROC) plot and the area under the ROC curve (AUC) describe how an adjustable threshold causes changes in two types of error: false positives and false negatives. Only part of the ROC curve and AUC are informative however when they are used with imbalanced data. Hence, alternatives to the AUC have been proposed, such as the partial AUC and the area under the precision-recall curve. However, these alternatives cannot be as fully interpreted as the AUC, in part because they ignore some information about actual negatives.Methods: We derive and propose a new concordant partial AUC and a new partial c statistic for ROC data-as foundational measures and methods to help understand and explain parts of the ROC plot and AUC. Our partial measures are continuous and discrete versions of the same measure, are derived from the AUC and c statistic respectively, are validated as equal to each other, and validated as equal in summation to whole measures where expected. Our partial measures are tested for validity on a classic ROC example from Fawcett, a variation thereof, and two real-life benchmark data sets in breast cancer: the Wisconsin and Ljubljana data sets. Interpretation of an example is then provided.Results: Results show the expected equalities between our new partial measures and the existing whole measures. The example interpretation illustrates the need for our newly derived partial measures.Conclusions: The concordant partial area under the ROC curve was proposed and unlike previous partial measure alternatives, it maintains the characteristics of the AUC. The first partial c statistic for ROC plots was also proposed as an unbiased interpretation for part of an ROC curve. The expected equalities among and between our newly derived partial measures and their existing full measure counterparts are confirmed. These measures may be used with any data set but this paper focuses on imbalanced data with low prevalence.Future Work: Future work with our proposed measures may: demonstrate their value for imbalanced data with high prevalence, compare them to other measures not based on areas; and combine them with other ROC measures and techniques. [ABSTRACT FROM AUTHOR]

Nowadays, breast cancer is considered one of the most common causes of death among adult women. At the same time, the bright side is that among all the types of cancer, breast cancer is more curable, if diagnosed in the early stages. In this paper, the diagnosis of breast cancer has been proposed using the least possible number of features based on correlation. In the proposed method, we have used correlation to find the strength between the input and the target features. Then we provided a way to create a new subset that consists of only the most relevant features. We have used the Wisconsin breast cancer data set (WBCD) for the experiments. The performance of the model is justified using classification accuracy and the f-score. The result shows that our proposed method obtained the highest classification accuracy (95.26%) with the Random Forest classification using only 4 features from 29 available features, which led to a reduction of 86% in data set size. [ABSTRACT FROM AUTHOR]

Missing Data (MD) is a common drawback that affects breast cancer classification. Thus, handling missing data is primordial before building any breast cancer classifier. This paper presents the impact of using ensemble Case-Based Reasoning (CBR) imputation on breast cancer classification. Thereafter, we evaluated the influence of CBR using parameter tuning and ensemble CBR (E-CBR) with three missingness mechanisms (MCAR: missing completely at random, MAR: missing at random and NMAR: not missing at random) and nine percentages (10% to 90%) on the accuracy rates of five classifiers: Decision trees, Random forest, K-nearest neighbor, Support vector machine and Multi-layer perceptron over two Wisconsin breast cancer datasets. All experiments were implemented using Weka JAVA API code 3.8; SPSS v20 was used for statistical tests. The findings confirmed that E-CBR yields to better results compared to CBR for the five classifiers. The MD percentage affects negatively the classifier performance: as the MD percentage increases, the accuracy rates of the classifier decrease regardless the MD mechanism and technique. RF with E-CBR outperformed all the other combinations (MD technique, classifier) with 89.72% for MCAR, 87.08% for MAR and 86.84% for NMAR. [ABSTRACT FROM AUTHOR]

Breast cancer (BC) is the largest cause of death in women. Accurate classification of breast cancer data is important in cancer diagnosis and classification of Malignant and Benign tumors can prevent patients to take unnecessary tests. Breast cancer classification can also be used to determine suitable treatment. Classification of Benign and Malignant patient groups is widely recognized research in the medical field. Due to the advantage of detecting critical features from a medical data set, machine learning is widely used in Breast cancer Prediction. Recently there has been greater attention to the use of machine learning methods in medical diagnosis. These decision support systems are effective and helpful for medical experts in the healthcare domain. The objective of this work is to address the problem of the classification of breast cancer data using ensemble learning. Ensemble learning techniques are used to improve the performance of a classifier. This paper deals with building a decision support system using the ensemble model built with Bayesian network and Radial Basis Function. In this work, extensive experiments were carried out on the much-studied open access data set "Wisconsin Breast Cancer Data set (WBCD)". The data set is partitioned into training and testing. Various metrics like accuracy, sensitivity, specificity, positive predictive value, negative predicted value, Error rate, false-positive rate, Mathew's correlation coefficient were used to measure the performance of the model. Experimental results show that the proposed method records a remarkable accuracy of 97% to classify breast cancer data and outperformed the existing approaches. The proposed ensemble learning would be viable in helping cancer specialists in recognizing cancer tumors accurately and help the patients in taking the correct treatment. [ABSTRACT FROM AUTHOR]

Objective: Medical data sets are usually small and have very high dimensionality. Too many attributes will make the analysis less efficient and will not necessarily increase accuracy, while too few data will decrease the modeling stability. Consequently, the main objective of this study is to extract the optimal subset of features to increase analytical performance when the data set is small., Methods: This paper proposes a fuzzy-based non-linear transformation method to extend classification related information from the original data attribute values for a small data set. Based on the new transformed data set, this study applies principal component analysis (PCA) to extract the optimal subset of features. Finally, we use the transformed data with these optimal features as the input data for a learning tool, a support vector machine (SVM). Six medical data sets: Pima Indians' diabetes, Wisconsin diagnostic breast cancer, Parkinson disease, echocardiogram, BUPA liver disorders dataset, and bladder cancer cases in Taiwan, are employed to illustrate the approach presented in this paper., Results: This research uses the t-test to evaluate the classification accuracy for a single data set; and uses the Friedman test to show the proposed method is better than other methods over the multiple data sets. The experiment results indicate that the proposed method has better classification performance than either PCA or kernel principal component analysis (KPCA) when the data set is small, and suggest creating new purpose-related information to improve the analysis performance., Conclusion: This paper has shown that feature extraction is important as a function of feature selection for efficient data analysis. When the data set is small, using the fuzzy-based transformation method presented in this work to increase the information available produces better results than the PCA and KPCA approaches., (Copyright © 2011 Elsevier B.V. All rights reserved.)

This paper introduces a novel classification rule mining model based on Pareto-based Multiobjective Optimization called CRM-PM. The process of rule extraction is a challenging classification task in data mining since it has several constraints and conflicting objectives such as accuracy and comprehensibility. In this study, this task is accepted as a multi-objective optimization problem. Classification accuracy and misclassification ratio are assigned as evaluation criteria. The candidate solutions are generated in the direction of a proposed strategy to determine optimal ranges of the attributes that form the rules. The proposed approach is applied on eight benchmark datasets (Iris Plants, Wine Quality, Glass Identification, Stat log (Heart), Haberman's Survival, E-coli, Wisconsin Breast Cancer, and Pima Indians Diabetes) included in the University of California at Irvine machine learning repository. Furthermore, CRM-PM is run in three different validation modes: cross-validation, training without test data, and training with random splitting. Regarding experimental results, it can be said that the presented method has a promising capability for classification, and it achieves comparative or superior results. [Display omitted] • A novel CRM model based on multiobjective optimization, run fast, and has a high accuracy rate. • A stepwise rule extraction process. • A new flexible structure to determine the optimum ranges of each attribute in multi-class data sets. • A new running model depends on a removing strategy that evolves the classification accuracy. • A fair comparison with 3 different validations: cross-validation, training without test, and random split. [ABSTRACT FROM AUTHOR]

Background: An individual's occupational activity (OA) may contribute significantly to daily physical activity (PA) and sedentary behavior (SB). However, there is little consensus about which occupational categories involve high OA or low OA, and the majority of categories are unclassifiable with current methods. The purpose of this study was to present population estimates of accelerometer-derived PA and SB variables for adults (n = 1112, 20-60 years) working the 40 occupational categories collected during the 2003-2004 National Health and Nutrition Examination Survey (NHANES). Methods: ActiGraph accelerometer-derived total activity counts/day (TAC), activity counts/minute, and proportion of wear time spent in moderate-to-vigorous PA [MVPA], lifestyle, and light PA organized by occupational category were ranked in ascending order and SB was ranked in descending order. Summing the ranks of the six accelerometer-derived variables generated a summary score for each occupational category, which was re-ranked in ascending order. Higher rankings indicated higher levels of OA, lower rankings indicated lower levels of OA. Tertiles of the summary score were used to establish three mutually exclusive accelerometer-determined OA groupings: high OA, intermediate OA, and low OA. Results: According to their summary score, 'farm and nursery workers' were classified as high OA and 'secretaries, stenographers, and typists' were classified as low OA. Consistent with previous research, some low OA occupational categories (e.g., 'engineers, architects, and scientists', 'technicians and related support occupations', 'management related occupations', 'executives, administrators, and managers', 'protective services', and 'writers, artists, entertainers, and athletes') associated with higher education and income had relatively greater amounts of MVPA compared to other low OA occupational categories, likely due to the greater percentage of men in those occupations and/or the influence of higher levels of leisure time PA. Men had more TAC, activity counts/minute and time in MVPA, but similar proportions of SB compared to women in all three OA groupings. Conclusions: Objectively measured PA allowed for a more precise estimate of the amount of PA and SB associated with different occupations and facilitated systematic classification of the 40 different occupational categories into three distinct OA groupings. This information provides new opportunities to explore the relationship between OA and health outcomes. [ABSTRACT FROM AUTHOR]

The problem with training spiking neural networks (SNNs) is relevant due to the ultra-low power consumption these networks could exhibit when implemented in neuromorphic hardware. The ongoing progress in the fabrication of memristors, a prospective basis for analogue synapses, gives relevance to studying the possibility of SNN learning on the base of synaptic plasticity models, obtained by fitting the experimental measurements of the memristor conductance change. The dynamics of memristor conductances is (necessarily) nonlinear, because conductance changes depend on the spike timings, which neurons emit in an all-or-none fashion. The ability to solve classification tasks was previously shown for spiking network models based on the bio-inspired local learning mechanism of spike-timing-dependent plasticity (STDP), as well as with the plasticity that models the conductance change of nanocomposite (NC) memristors. Input data were presented to the network encoded into the intensities of Poisson input spike sequences. This work considers another approach for encoding input data into input spike sequences presented to the network: temporal encoding, in which an input vector is transformed into relative timing of individual input spikes. Since temporal encoding uses fewer input spikes, the processing of each input vector by the network can be faster and more energy-efficient. The aim of the current work is to show the applicability of temporal encoding to training spiking networks with three synaptic plasticity models: STDP, NC memristor approximation, and PPX memristor approximation. We assess the accuracy of the proposed approach on several benchmark classification tasks: Fisher's Iris, Wisconsin breast cancer, and the pole balancing task (CartPole). The accuracies achieved by SNN with memristor plasticity and conventional STDP are comparable and are on par with classic machine learning approaches. [ABSTRACT FROM AUTHOR]

Interest in understanding physical and hydraulic factors that might drive distribution and abundance of freshwater mussels has been increasing due to their decline throughout North America. We assessed whether the spatial distribution of unionid mussels could be predicted from physical and hydraulic variables in a reach of the Upper Mississippi River. Classification and regression tree (CART) models were constructed using mussel data compiled from various sources and explanatory variables derived from GIS coverages. Prediction success of CART models for presence–absence of mussels ranged from 71 to 76% across three gears (brail, sled-dredge, and dive-quadrat) and 51% of the deviance in abundance. Models were largely driven by shear stress and substrate stability variables, but interactions with simple physical variables, especially slope, were also important. Geospatial models, which were based on tree model results, predicted few mussels in poorly connected backwater areas (e.g., floodplain lakes) and the navigation channel, whereas main channel border areas with high geomorphic complexity (e.g., river bends, islands, side channel entrances) and small side channels were typically favorable to mussels. Moreover, bootstrap aggregation of discharge-specific regression tree models of dive-quadrat data indicated that variables measured at low discharge were about 25% more predictive (PMSE = 14.8) than variables measured at median discharge (PMSE = 20.4) with high discharge (PMSE = 17.1) variables intermediate. This result suggests that episodic events such as droughts and floods were important in structuring mussel distributions. Although the substantial mussel and ancillary data in our study reach is unusual, our approach to develop exploratory statistical and geospatial models should be useful even when data are more limited. [ABSTRACT FROM AUTHOR]

Two Woodfordian glaciated regions in Wisconsin were analyzed to determine the impact of variable loess thickness and contrasting till texture on soil pattern and landscape properties. Each region consisted of a ground moraine/end moraine/outwash plain landform sequence. Data were compiled from soil and topographic maps. In addition to using standard methods of soilscape analysis. I used a natural soil drainage index, calculated from taxonomic data, to determine regional variability and pattern of soil wetness. Results indicate that soilscape variation within each region was large. There were no statistically significant differences, therefore, that could be attributed to known soil- forming processes or to parent material variability. Stratifying the data by geomorphic subregion rather than by region reduced the within-unit soil variability and resulted in numerous significant soilscape differences between subregions. The implication is that soilscape variability is large at the regional scale but converges rapidly with decreasing size of study area. [ABSTRACT FROM AUTHOR]

Background: As of today, cancer is still one of the most prevalent and high-mortality diseases, summing more than 9 million deaths in 2018. This has motivated researchers to study the application of machine learning-based solutions for cancer detection to accelerate its diagnosis and help its prevention. Among several approaches, one is to automatically classify tumor samples through their gene expression analysis.Methods: In this work, we aim to distinguish five different types of cancer through RNA-Seq datasets: thyroid, skin, stomach, breast, and lung. To do so, we have adopted a previously described methodology, with which we compare the performance of 3 different autoencoders (AEs) used as a deep neural network weight initialization technique. Our experiments consist in assessing two different approaches when training the classification model - fixing the weights after pre-training the AEs, or allowing fine-tuning of the entire network - and two different strategies for embedding the AEs into the classification network, namely by only importing the encoding layers, or by inserting the complete AE. We then study how varying the number of layers in the first strategy, the AEs latent vector dimension, and the imputation technique in the data preprocessing step impacts the network's overall classification performance. Finally, with the goal of assessing how well does this pipeline generalize, we apply the same methodology to two additional datasets that include features extracted from images of malaria thin blood smears, and breast masses cell nuclei. We also discard the possibility of overfitting by using held-out test sets in the images datasets.Results: The methodology attained good overall results for both RNA-Seq and image extracted data. We outperformed the established baseline for all the considered datasets, achieving an average F1 score of 99.03, 89.95, and 98.84 and an MCC of 0.99, 0.84, and 0.98, for the RNA-Seq (when detecting thyroid cancer), the Malaria, and the Wisconsin Breast Cancer data, respectively.Conclusions: We observed that the approach of fine-tuning the weights of the top layers imported from the AE reached higher results, for all the presented experiences, and all the considered datasets. We outperformed all the previous reported results when comparing to the established baselines. [ABSTRACT FROM AUTHOR]