Jin Huang

Bio: Jin Huang is an academic researcher from University of Ottawa. The author has contributed to research in topics: Online machine learning & Ranking SVM. The author has an hindex of 6, co-authored 13 publications receiving 3155 citations. Previous affiliations of Jin Huang include University of Western Ontario.

Using AUC and accuracy in evaluating learning algorithms

Abstract: The area under the ROC (receiver operating characteristics) curve, or simply AUC, has been traditionally used in medical diagnosis since the 1970s. It has recently been proposed as an alternative single-number measure for evaluating the predictive ability of learning algorithms. However, no formal arguments were given as to why AUC should be preferred over accuracy. We establish formal criteria for comparing two different measures for learning algorithms and we show theoretically and empirically that AUC is a better measure (defined precisely) than accuracy. We then reevaluate well-established claims in machine learning based on accuracy using AUC and obtain interesting and surprising new results. For example, it has been well-established and accepted that Naive Bayes and decision trees are very similar in predictive accuracy. We show, however, that Naive Bayes is significantly better than decision trees in AUC. The conclusions drawn in this paper may make a significant impact on machine learning and data mining applications.

AUC: a statistically consistent and more discriminating measure than accuracy

Abstract: Predictive accuracy has been used as the main and often only evaluation criterion for the predictive performance of classification learning algorithms. In recent years, the area under the ROC (Receiver Operating Characteristics) curve, or simply AUC, has been proposed as an alternative single-number measure for evaluating learning algorithms. In this paper, we prove that AUC is a better measure than accuracy. More specifically, we present rigourous definitions on consistency and discriminancy in comparing two evaluation measures for learning algorithms. We then present empirical evaluations and a formal proof to establish that AUC is indeed statistically consistent and more discriminating than accuracy. Our result is quite significant since we formally prove that, for the first time, AUC is a better measure than accuracy in the evaluation of learning algorithms.

AUC: a better measure than accuracy in comparing learning algorithms

Abstract: Predictive accuracy has been widely used as the main criterion for comparing the predictive ability of classification systems (such as C4.5, neural networks, and Naive Bayes). Most of these classifiers also produce probability estimations of the classification, but they are completely ignored in the accuracy measure. This is often taken for granted because both training and testing sets only provide class labels. In this paper we establish rigourously that, even in this setting, the area under the ROC (Receiver Operating Characteristics) curve, or simply AUC, provides a better measure than accuracy. Our result is quite significant for three reasons. First, we establish, for the first time, rigourous criteria for comparing evaluation measures for learning algorithms. Second, it suggests that AUC should replace accuracy when measuring and comparing classification systems. Third, our result also prompts us to reevaluate many well-established conclusions based on accuracy in machine learning. For example, it is well accepted in the machine learning community that, in terms of predictive accuracy, Naive Bayes and decision trees are very similar. Using AUC, however, we show experimentally that Naive Bayes is significantly better than the decision-tree learning algorithms.

Comparing naive Bayes, decision trees, and SVM with AUC and accuracy

Abstract: Predictive accuracy has often been used as the main and often only evaluation criterion for the predictive performance of classification or data mining algorithms. In recent years, the area under the ROC (receiver operating characteristics) curve, or simply AUC, has been proposed as an alternative single-number measure for evaluating performance of learning algorithms. We proved that AUC is, in general, a better measure (defined precisely) than accuracy. Many popular data mining algorithms should then be reevaluated in terms of AUC. For example, it is well accepted that Naive Bayes and decision trees are very similar in accuracy. How do they compare in AUC? Also, how does the recently developed SVM (support vector machine) compare to traditional learning algorithms in accuracy and AUC? We will answer these questions. Our conclusions will provide important guidelines in data mining applications on real-world datasets.

A systematic analysis of performance measures for classification tasks

Abstract: This paper presents a systematic analysis of twenty four performance measures used in the complete spectrum of Machine Learning classification tasks, i.e., binary, multi-class, multi-labelled, and hierarchical. For each classification task, the study relates a set of changes in a confusion matrix to specific characteristics of data. Then the analysis concentrates on the type of changes to a confusion matrix that do not change a measure, therefore, preserve a classifier's evaluation (measure invariance). The result is the measure invariance taxonomy with respect to all relevant label distribution changes in a classification problem. This formal analysis is supported by examples of applications where invariance properties of measures lead to a more reliable evaluation of classifiers. Text classification supplements the discussion with several case studies.

The advantages of the Matthews correlation coefficient (MCC) over F1 score and accuracy in binary classification evaluation

Abstract: To evaluate binary classifications and their confusion matrices, scientific researchers can employ several statistical rates, accordingly to the goal of the experiment they are investigating. Despite being a crucial issue in machine learning, no widespread consensus has been reached on a unified elective chosen measure yet. Accuracy and F1 score computed on confusion matrices have been (and still are) among the most popular adopted metrics in binary classification tasks. However, these statistical measures can dangerously show overoptimistic inflated results, especially on imbalanced datasets. The Matthews correlation coefficient (MCC), instead, is a more reliable statistical rate which produces a high score only if the prediction obtained good results in all of the four confusion matrix categories (true positives, false negatives, true negatives, and false positives), proportionally both to the size of positive elements and the size of negative elements in the dataset. In this article, we show how MCC produces a more informative and truthful score in evaluating binary classifications than accuracy and F1 score, by first explaining the mathematical properties, and then the asset of MCC in six synthetic use cases and in a real genomics scenario. We believe that the Matthews correlation coefficient should be preferred to accuracy and F1 score in evaluating binary classification tasks by all scientific communities.

A Review on Ensembles for the Class Imbalance Problem: Bagging-, Boosting-, and Hybrid-Based Approaches

Abstract: Classifier learning with data-sets that suffer from imbalanced class distributions is a challenging problem in data mining community. This issue occurs when the number of examples that represent one class is much lower than the ones of the other classes. Its presence in many real-world applications has brought along a growth of attention from researchers. In machine learning, the ensemble of classifiers are known to increase the accuracy of single classifiers by combining several of them, but neither of these learning techniques alone solve the class imbalance problem, to deal with this issue the ensemble learning algorithms have to be designed specifically. In this paper, our aim is to review the state of the art on ensemble techniques in the framework of imbalanced data-sets, with focus on two-class problems. We propose a taxonomy for ensemble-based methods to address the class imbalance where each proposal can be categorized depending on the inner ensemble methodology in which it is based. In addition, we develop a thorough empirical comparison by the consideration of the most significant published approaches, within the families of the taxonomy proposed, to show whether any of them makes a difference. This comparison has shown the good behavior of the simplest approaches which combine random undersampling techniques with bagging or boosting ensembles. In addition, the positive synergy between sampling techniques and bagging has stood out. Furthermore, our results show empirically that ensemble-based algorithms are worthwhile since they outperform the mere use of preprocessing techniques before learning the classifier, therefore justifying the increase of complexity by means of a significant enhancement of the results.

Fake News Detection on Social Media: A Data Mining Perspective

Abstract: Social media for news consumption is a double-edged sword. On the one hand, its low cost, easy access, and rapid dissemination of information lead people to seek out and consume news from social media. On the other hand, it enables the wide spread of \fake news", i.e., low quality news with intentionally false information. The extensive spread of fake news has the potential for extremely negative impacts on individuals and society. Therefore, fake news detection on social media has recently become an emerging research that is attracting tremendous attention. Fake news detection on social media presents unique characteristics and challenges that make existing detection algorithms from traditional news media ine ective or not applicable. First, fake news is intentionally written to mislead readers to believe false information, which makes it difficult and nontrivial to detect based on news content; therefore, we need to include auxiliary information, such as user social engagements on social media, to help make a determination. Second, exploiting this auxiliary information is challenging in and of itself as users' social engagements with fake news produce data that is big, incomplete, unstructured, and noisy. Because the issue of fake news detection on social media is both challenging and relevant, we conducted this survey to further facilitate research on the problem. In this survey, we present a comprehensive review of detecting fake news on social media, including fake news characterizations on psychology and social theories, existing algorithms from a data mining perspective, evaluation metrics and representative datasets. We also discuss related research areas, open problems, and future research directions for fake news detection on social media.

A systematic study of the class imbalance problem in convolutional neural networks

Abstract: In this study, we systematically investigate the impact of class imbalance on classification performance of convolutional neural networks (CNNs) and compare frequently used methods to address the issue. Class imbalance is a common problem that has been comprehensively studied in classical machine learning, yet very limited systematic research is available in the context of deep learning. In our study, we use three benchmark datasets of increasing complexity, MNIST, CIFAR-10 and ImageNet, to investigate the effects of imbalance on classification and perform an extensive comparison of several methods to address the issue: oversampling, undersampling, two-phase training, and thresholding that compensates for prior class probabilities. Our main evaluation metric is area under the receiver operating characteristic curve (ROC AUC) adjusted to multi-class tasks since overall accuracy metric is associated with notable difficulties in the context of imbalanced data. Based on results from our experiments we conclude that (i) the effect of class imbalance on classification performance is detrimental; (ii) the method of addressing class imbalance that emerged as dominant in almost all analyzed scenarios was oversampling; (iii) oversampling should be applied to the level that completely eliminates the imbalance, whereas the optimal undersampling ratio depends on the extent of imbalance; (iv) as opposed to some classical machine learning models, oversampling does not cause overfitting of CNNs; (v) thresholding should be applied to compensate for prior class probabilities when overall number of properly classified cases is of interest.