The general approach taken when training deep learning classifiers is to save the parameters after every few iterations, train until either a human observer or a simple metric-based heuristic decides the network isn't learning anymore, and then backtrack and pick the saved parameters with the best validation accuracy. Simple methods are used to determine if a neural network isn't learning anymore because, as long as it's well after the optimal values are found, the condition doesn't impact the final accuracy of the model. However from a runtime perspective, this is of great significance to the many cases where numerous neural networks are trained simultaneously (e.g. hyper-parameter tuning). Motivated by this, we introduce a statistical significance test to determine if a neural network has stopped learning. This stopping criterion appears to represent a happy medium compared to other popular stopping criterions, achieving comparable accuracy to the criterions that achieve the highest final accuracies in 77% or fewer epochs, while the criterions which stop sooner do so with an appreciable loss to final accuracy. Additionally, we use this as the basis of a new learning rate scheduler, removing the need to manually choose learning rate schedules and acting as a quasi-line search, achieving superior or comparable empirical performance to existing methods.
Finding statistically significant interactions between binary variables is computationally and statistically challenging in high-dimensional settings, due to the combinatorial explosion in the number of hypotheses. Terada et al. recently showed how to elegantly address this multiple testing problem by excluding non-testable hypotheses. Still, it remains unclear how their approach scales to large datasets. We here proposed strategies to speed up the approach by Terada et al. and evaluate them thoroughly in 11 real-world benchmark datasets. We observe that one approach, incremental search with e…
GradES: Significantly Faster Training in Transformers with Gradient-Based Early Stopping
Department of Computer Science, Boston University Metropolitan College（波士顿大学计算机科学系）；Information Hub, The Hong Kong University of Science and Technology, Guangzhou（香港科技大学广州信息中心）；School of Engineering and Technology, Duy Tan University, Da Nang, Vietnam（杜益大学工程科技学院，岘港，越南）；Department of AI, School of Computer Science and Engineering, Galgotias University, Greater Noida, India（加洛吉亚大学人工智能系，诺伊达，印度）
Technical University of Munich（慕尼黑工业大学）；Rice University（莱斯大学）
Over-parameterized models, such as large deep networks, often exhibit a double descent phenomenon, whereas a function of model size, error first decreases, increases, and decreases at last. This intriguing double descent behavior also occurs as a function of training epochs and has been conjectured to arise because training epochs control the model complexity. In this paper, we show that such epoch-wise double descent arises for a different reason: It is caused by a superposition of two or more bias-variance tradeoffs that arise because different parts of the network are learned at different e…
Toyota Technological Institute at Chicago（芝加哥丰田理工学院）；Purdue University（普渡大学）
We propose a Las Vegas transformation of Markov Chain Monte Carlo (MCMC) estimators of Restricted Boltzmann Machines (RBMs). We denote our approach Markov Chain Las Vegas (MCLV). MCLV gives statistical guarantees in exchange for random running times. MCLV uses a stopping set built from the training data and has maximum number of Markov chain steps K (referred as MCLV-K). We present a MCLV-K gradient estimator (LVS-K) for RBMs and explore the correspondence and differences between LVS-K and Contrastive Divergence (CD-K), with LVS-K significantly outperforming CD-K training RBMs over the MNIST d…
Graduate School of Knowledge Service Engineering（知识服务工程研究生院）
Noisy labels are very common in real-world training data, which lead to poor generalization on test data because of overfitting to the noisy labels. In this paper, we claim that such overfitting can be avoided by "early stopping" training a deep neural network before the noisy labels are severely memorized. Then, we resume training the early stopped network using a "maximal safe set," which maintains a collection of almost certainly true-labeled samples at each epoch since the early stop point. Putting them all together, our novel two-phase training method, called Prestopping, realizes noise-f…
Universitat Politècnica de Catalunya（加泰罗尼亚理工大学）；University of Southern California（南加州大学）
State-of-the-art neural network architectures continue to scale in size and deliver impressive generalization results, although this comes at the expense of limited interpretability. In particular, a key challenge is to determine when to stop training the model, as this has a significant impact on generalization. Convolutional neural networks (ConvNets) comprise high-dimensional feature spaces formed by the aggregation of multiple channels, where analyzing intermediate data representations and the model's evolution can be challenging owing to the curse of dimensionality. We present channel-wis…
National Institute of Informatics（国立信息学研究所）；JST PRESTO（日本科学技术振兴机构PRESTO）；ETH Zürich（苏黎世联邦理工学院）；SIB Swiss Institute of Bioinformatics（SIB瑞士生物信息学研究所）
The search for higher-order feature interactions that are statistically significantly associated with a class variable is of high relevance in fields such as Genetics or Healthcare, but the combinatorial explosion of the candidate space makes this problem extremely challenging in terms of computational efficiency and proper correction for multiple testing. While recent progress has been made regarding this challenge for binary features, we here present the first solution for continuous features. We propose an algorithm which overcomes the combinatorial explosion of the search space of higher-o…
We propose a new computational approach for tracking and detecting statistically significant linguistic shifts in the meaning and usage of words. Such linguistic shifts are especially prevalent on the Internet, where the rapid exchange of ideas can quickly change a word's meaning. Our meta-analysis approach constructs property time series of word usage, and then uses statistically sound change point detection algorithms to identify significant linguistic shifts. We consider and analyze three approaches of increasing complexity to generate such linguistic property time series, the culmination o…
NOVA IMS, Universidade Nova de Lisboa（新里斯本大学NOVA IMS）；Faculty of Sciences, University of Lisbon（里斯本大学理学院）；University of Coimbra（科英布拉大学）
In iterative supervised learning algorithms it is common to reach a point in the search where no further induction seems to be possible with the available data. If the search is continued beyond this point, the risk of overfitting increases significantly. Following the recent developments in inductive semantic stochastic methods, this paper studies the feasibility of using information gathered from the semantic neighborhood to decide when to stop the search. Two semantic stopping criteria are proposed and experimentally assessed in Geometric Semantic Genetic Programming (GSGP) and in the Seman…
Discriminative pattern mining is an essential task of data mining. This task aims to discover patterns which occur more frequently in a class than other classes in a class-labeled dataset. This type of patterns is valuable in various domains such as bioinformatics, data classification. In this paper, we propose a novel algorithm, named SSDPS, to discover patterns in two-class datasets. The SSDPS algorithm owes its efficiency to an original enumeration strategy of the patterns, which allows to exploit some degrees of anti-monotonicity on the measures of discriminance and statistical significanc…
Systems for ML inference are widely deployed today, but they typically optimize ML inference workloads using techniques designed for conventional data serving workloads and miss critical opportunities to leverage the statistical nature of ML. In this paper, we present Willump, an optimizer for ML inference that introduces two statistically-motivated optimizations targeting ML applications whose performance bottleneck is feature computation. First, Willump automatically cascades feature computation for classification queries: Willump classifies most data inputs using only high-value, low-cost f…
Finnish Institute of Occupational Health（芬兰职业健康研究所）
In many data exploration tasks it is meaningful to identify groups of attribute interactions that are specific to a variable of interest. For instance, in a dataset where the attributes are medical markers and the variable of interest (class variable) is binary indicating presence/absence of disease, we would like to know which medical markers interact with respect to the binary class label. These interactions are useful in several practical applications, for example, to gain insight into the structure of the data, in feature selection, and in data anonymisation. We present a novel method, bas…
