Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userSurvey Equivalence: A Procedure for Measuring Classifier Accuracy Against Human Labels
83
0
0.0
(
0
)
Added by
Tim Weninger PhD
Publication date
2021
fields
Informatics Engineering
and research's language is
English
Ask ChatGPT about the research
No Arabic abstract
In many classification tasks, the ground truth is either noisy or subjective. Examples include: which of two alternative paper titles is better? is this comment toxic? what is the political leaning of this news article? We refer to such tasks as survey settings because the ground truth is defined through a survey of one or more human raters. In survey settings, conventional measurements of classifier accuracy such as precision, recall, and cross-entropy confound the quality of the classifier with the level of agreement among human raters. Thus, they have no meaningful interpretation on their own. We describe a procedure that, given a dataset with predictions from a classifier and K ratings per item, rescales any accuracy measure into one that has an intuitive interpretation. The key insight is to score the classifier not against the best proxy for the ground truth, such as a majority vote of the raters, but against a single human rater at a time. That score can be compared to other predictors scores, in particular predictors created by combining labels from several other human raters. The survey equivalence of any classifier is the minimum number of raters needed to produce the same expected score as that found for the classifier.
rate research
Read More
To calculate the model accuracy on a computer vision task, e.g., object recognition, we usually require a test set composing of test samples and their ground truth labels. Whilst standard usage cases satisfy this requirement, many real-world scenarios involve unlabeled test data, rendering common model evaluation methods infeasible. We investigate this important and under-explored problem, Automatic model Evaluation (AutoEval). Specifically, given a labeled training set and a classifier, we aim to estimate the classification accuracy on unlabeled test datasets. We construct a meta-dataset: a dataset comprised of datasets generated from the original images via various transformations such as rotation, background substitution, foreground scaling, etc. As the classification accuracy of the model on each sample (dataset) is known from the original dataset labels, our task can be solved via regression. Using the feature statistics to represent the distribution of a sample dataset, we can train regression models (e.g., a regression neural network) to predict model performance. Using synthetic meta-dataset and real-world datasets in training and testing, respectively, we report a reasonable and promising prediction of the model accuracy. We also provide insights into the application scope, limitation, and potential future direction of AutoEval.
Motivated by an open problem of validating protein identities in label-free shotgun proteomics work-flows, we present a testing procedure to validate class/protein labels using available measurements across instances/peptides. More generally, we present a solution to the problem of identifying instances that are deemed, based on some distance (or quasi-distance) measure, as outliers relative to the subset of instances assigned to the same class. The proposed procedure is non-parametric and requires no specific distributional assumption on the measured distances. The only assumption underlying the testing procedure is that measured distances between instances within the same class are stochastically smaller than measured distances between instances from different classes. The test is shown to simultaneously control the Type I and Type II error probabilities whilst also controlling the overall error probability of the repeated testing invoked in the validation procedure of initial class labeling. The theoretical results are supplemented with results from an extensive numerical study, simulating a typical setup for labeling validation in proteomics work-flow applications. These results illustrate the applicability and viability of our method. Even with up to 25% of instances mislabeled, our testing procedure maintains a high specificity and greatly reduces the proportion of mislabeled instances.
Modern neural networks have the capacity to overfit noisy labels frequently found in real-world datasets. Although great progress has been made, existing techniques are limited in providing theoretical guarantees for the performance of the neural networks trained with noisy labels. Here we propose a novel approach with strong theoretical guarantees for robust training of deep networks trained with noisy labels. The key idea behind our method is to select weighted subsets (coresets) of clean data points that provide an approximately low-rank Jacobian matrix. We then prove that gradient descent applied to the subsets do not overfit the noisy labels. Our extensive experiments corroborate our theory and demonstrate that deep networks trained on our subsets achieve a significantly superior performance compared to state-of-the art, e.g., 6% increase in accuracy on CIFAR-10 with 80% noisy labels, and 7% increase in accuracy on mini Webvision.
Todays available datasets in the wild, e.g., from social media and open platforms, present tremendous opportunities and challenges for deep learning, as there is a significant portion of tagged images, but often with noisy, i.e. erroneous, labels. Recent studies improve the robustness of deep models against noisy labels without the knowledge of true labels. In this paper, we advocate to derive a stronger classifier which proactively makes use of the noisy labels in addition to the original images - turning noisy labels into learning features. To such an end, we propose a novel framework, ExpertNet, composed of Amateur and Expert, which iteratively learn from each other. Amateur is a regular image classifier trained by the feedback of Expert, which imitates how human experts would correct the predicted labels from Amateur using the noise pattern learnt from the knowledge of both the noisy and ground truth labels. The trained Amateur and Expert proactively leverage the images and their noisy labels to infer image classes. Our empirical evaluations on noi
Semi-supervised node classification, as a fundamental problem in graph learning, leverages unlabeled nodes along with a small portion of labeled nodes for training. Existing methods rely heavily on high-quality labels, which, however, are expensive to obtain in real-world applications since certain noises are inevitably involved during the labeling process. It hence poses an unavoidable challenge for the learning algorithm to generalize well. In this paper, we propose a novel robust learning objective dubbed pairwise interactions (PI) for the model, such as Graph Neural Network (GNN) to combat noisy labels. Unlike classic robust training approaches that operate on the pointwise interactions between node and class label pairs, PI explicitly forces the embeddings for node pairs that hold a positive PI label to be close to each other, which can be applied to both labeled and unlabeled nodes. We design several instantiations for PI labels based on the graph structure and the node class labels, and further propose a new uncertainty-aware training technique to mitigate the negative effect of the sub-optimal PI labels. Extensive experiments on different datasets and GNN architectures demonstrate the effectiveness of PI, yielding a promising improvement over the state-of-the-art methods.
suggested questions
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
