Develops a framework to test excessive influence of small data subsets.
problem Identifying when small data subsets significantly impact model conclusions.
method Formalizes the concept of most influential sets, deriving influence formulas and extreme value distributions.
result Allows rigorous hypothesis testing for excessive influence, resolving contested findings.
New method estimates data influence efficiently by leveraging test samples.
problem Efficiently estimating influence of training data on model predictions.
method Mirrored Influence Hypothesis, forward pass for test samples.
result Significant improvement in efficiency over existing methods.
With the rapid adoption of machine learning systems in sensitive applications, there is an increasing need to make black-box models explainable. Often we want to identify an influential group of training samples in a particular test prediction for a given machine learning model. Existing influence functions tackle this…
We introduce kernel nonparametric tests for Lancaster three-variable interaction and for total independence, using embeddings of signed measures into a reproducing kernel Hilbert space. The resulting test statistics are straightforward to compute, and are used in powerful interaction tests, which are consistent against…
Improved scalability and interpretability in training data attribution.
problem Identifying which training data drives specific behaviors, especially unintended ones.
method Leveraging interpretable structures within the model to attribute model behavior to semantic directions, not individual test examples.
result Simple probe-based attribution methods are first-order approximations of Concept Influence that achieve comparable performance while being over an order-of-magnitude faster.
Influence functions are inaccurate in deep learning models, especially for deeper networks.
problem Inaccuracies in influence functions in deep learning models.
method Empirical study of influence functions in neural network models trained on various datasets.
result Influence estimates are often erroneous for deeper networks and require regularization.
Motivated by the need to audit complex and black box models, there has been extensive research on quantifying how data features influence model predictions. Feature influence can be direct (a direct influence on model outcomes) and indirect (model outcomes are influenced via proxy features). Feature influence can also …
AdaBoost's success explained through noise influence measure.
problem Understanding why AdaBoost is a successful classifier.
method Introduced a measure of noise influence (ION) to explain AdaBoost's success.
result ION decreases with iteration number and base learner complexity.
The paper extends influence functions to sequence tagging tasks for better model interpretability.
problem Lack of interpretability methods for sequence tagging models.
method Define and compute influence of training instance segments on test segment predictions.
result The segment influence method tracks with true influence and identifies annotation errors.
Paper introduces a multi-stage influence function to track model predictions.
problem Improving natural language processing and computer vision performance.
method Develops a multi-stage influence function to track predictions from finetuned models back to pretraining data.
result Identifies pretraining examples contributing most to finetuning task predictions.
In2Core selects a coreset for efficient LLM fine-tuning with reduced data.
problem Costly fine-tuning of large language models due to extensive parameters and data requirements.
method Analyzes model gradients to estimate training sample influence, optimizing for efficiency.
result Achieves similar performance with 50% of training data using In2Core.
New framework detects directional influence in multivariate time series.
problem Detecting directional influence in multivariate time series.
method Order-constrained spectral non-invariance.
result Unique diagnostic functional for directional influence.
We apply a wild bootstrap method to the Lancaster three-variable interaction measure in order to detect factorisation of the joint distribution on three variables forming a stationary random process, for which the existing permutation bootstrap method fails. As in the i.i.d. case, the Lancaster test is found to outperf…
We investigated the network structures of the Japanese stock market through the minimum spanning tree. We defined grouping coefficient to test the validity of conventional grouping by industrial categories, and found a decreasing in trend for the coefficient. This phenomenon supports the increasing external influences …
We develop a pivotal test to assess the statistical significance of the feature variables in a single-layer feedforward neural network regression model. We propose a gradient-based test statistic and study its asymptotics using nonparametric techniques. Under technical conditions, the limiting distribution is given by …
Causal influence measures for machine learnt classifiers shed light on the reasons behind classification, and aid in identifying influential input features and revealing their biases. However, such analyses involve evaluating the classifier using datapoints that may be atypical of its training distribution. Standard me…
Researchers study how teachers' advising relationships influence their perceptions of satisfaction and students, not policy influence.
problem Understanding the relationship between teachers' advising relationships and their perceptions of satisfaction and students.
method Proposed a novel joint model of network and item responses (JNIRM) with correlated latent variables.
result Teachers' advising relationships contribute more to satisfaction and students than to influence over educational policies.
In the time of Big Data, training complex models on large-scale data sets is challenging, making it appealing to reduce data volume for saving computation resources by subsampling. Most previous works in subsampling are weighted methods designed to help the performance of subset-model approach the full-set-model, hence…
Captures data influence changes during training.
problem Traditional influence functions fail for modern training methods.
method Formalized trajectory-specific LOO influence, using data value embedding.
result Data influence varies by training stage, early and late stages have greater impact.
Study predicts social relationships using triadic influence from social networks.
problem Difficulty in quantifying social relationships and their dynamics.
method Real social networks of 13 schools, neural networks, high-dimensional embedding.
result Triadic influence achieves highest accuracy in predicting student relationships.
Paper introduces Influence Function to assess OOD generalization stability.
problem Assessing OOD generalization accuracy when target domains are unknown.
method Introduced Influence Function from robust statistics to monitor model stability.
result Accuracy on test domains and Influence Function variance can distinguish OOD algorithms and generalization quality.
In recent years, probabilistic forecasts techniques were proposed in research as well as in applications to integrate volatile renewable energy resources into the electrical grid. These techniques allow decision makers to take the uncertainty of the prediction into account and, therefore, to devise optimal decisions, e…
Optimal intervention in economic networks modeled as influence maximization, with hard computational problems.
problem Optimal intervention in economic networks modeled as influence maximization.
method Transformed into influence maximization-like form, with theoretical and practical implications.
result Optimal intervention is NP-hard and cannot be approximated to a constant factor in polynomial time.
New experiments show deep networks benefit from memorizing rare data points.
problem Understanding why neural networks perform well on training data.
method Estimating influence and memorization values of training examples.
result Deep networks benefit from memorizing rare data points for good performance.
Current approaches for explaining machine learning models fall into two distinct classes: antecedent event influence and value attribution. The former leverages training instances to describe how much influence a training point exerts on a test point, while the latter attempts to attribute value to the features most pe…
Gradient rollback explains neural models by identifying influential training examples.
problem Explain predictions of neural black-box models, especially in applications requiring user trust.
method Gradient rollback, a general approach for influence estimation applicable to neural models with limited parameter updates.
result Gradient rollback provides faithful explanations for neural models, including knowledge graph embedding methods.
Testing whether a probability distribution is compatible with a given Bayesian network is a fundamental task in the field of causal inference, where Bayesian networks model causal relations. Here we consider the class of causal structures where all correlations between observed quantities are solely due to the influenc…
MRI image quality affects statistical and predictive analysis of brain morphology.
problem Impact of MRI image quality on statistical and predictive analysis of brain morphology.
method Systematic testing of image quality on univariate statistics and machine learning classification using three large datasets.
result Low-quality MRI data significantly affects detecting significant sex/gender differences in smaller samples, but not in larger ones.
New method controls false discoveries in online testing with deadlines.
problem Controlling false discoveries in online hypothesis testing with decision deadlines.
method Benjamini-Hochberg-type procedure over a moving window of hypotheses with adaptive threshold parameters.
result Controls false discovery rate at every stage and adaptively chosen stopping times.
GOIMDA selects inputs to maximize expected influence on a goal functional, reducing data acquisition needs.
problem Challenges in active data acquisition for learning and optimization tasks in deep neural networks.
method GOIMDA uses inverse curvature and goal gradient to select inputs maximizing expected influence on a specified goal functional.
result GOIMDA achieves target performance with fewer labeled samples or function evaluations compared to baselines.
New framework tests mean-variance spanning in high dimensions.
problem Testing mean-variance spanning in high-dimensional asset spaces.
method Robust Student-t statistic based on batch-mean method, combined using Cauchy combination test.
result Advantages of diversification vary by economic conditions and cross-country.
Study explores factors influencing saving behavior among Dhaka employees.
problem Factors influencing saving behavior among Dhaka employees.
method Quantitative approach with cross-sectional survey design, structured questionnaire, descriptive statistics, reliability analysis, regression analysis.
result Only financial management practices had a significant positive relationship with saving behavior.
Consistently checking the statistical significance of experimental results is one of the mandatory methodological steps to address the so-called "reproducibility crisis" in deep reinforcement learning. In this tutorial paper, we explain how the number of random seeds relates to the probabilities of statistical errors. …
This work develops a non-parametric test for relational independence in non-i.i.d. data.
problem Testing independence in relational systems where data samples are not i.i.d.
method Kernel mean embedding for relational variables, consistent non-parametric scalable kernel test.
result Empirically validated effectiveness compared to state-of-the-art tests.
TracIn measures how training examples affect model predictions.
problem Measuring the influence of training examples on model predictions.
method Tracing changes in loss during training when a specific example is used.
result TracIn provides a scalable method to estimate example influence.
Study uses Granger causality to show investor sentiment influences stock prices.
problem Understanding the relationship between investor sentiment and stock market movements.
method Applied Granger causality to analyze the relationship between close price index and sentiment score.
result Sentiment analysis shows a positive correlation with stock price movements.
GTBO uses group testing to optimize high-dimensional functions efficiently.
problem Challenges in optimizing high-dimensional, expensive functions due to the curse of dimensionality.
method GTBO combines testing and optimization phases to identify active variables and guide efficient optimization.
result GTBO outperforms state-of-the-art methods on high-dimensional optimization tasks.
We propose a mathematical procedure for finding informed traders in ultra-high frequency trading. We wrote it as Vector ARMA and found condition of its stationarity. For the price exposure complied with ARMA(1,2) we proved that underlying asset price difference can be derived as ARMA(1,1) process. For validation of the…
Many widely studied graphical models with latent variables lead to nontrivial constraints on the distribution of the observed variables. Inspired by the Bell inequalities in quantum mechanics, we refer to any linear inequality whose violation rules out some latent variable model as a "hidden variable test" for that mod…
New term ADS describes how machine learning can change user behavior.
problem Machine learning systems can unintentionally change user behavior, affecting performance.
method Introduced `unit tests` and mitigation strategy for hidden incentives in auto-induced distributional shift.
result Meta-learning and Q-learning sometimes fail unit tests but pass with mitigation strategy.
Framework analyzes stock price co-movement with fundamentals using big data.
problem Understanding complex relationships between stock price co-movements and fundamental characteristics.
method Advanced big data techniques, four regression models.
result Identifies leading co-movement stocks and their influencing factors.
Religious adherence reduces corporate greenwashing behavior.
problem Greenwashing behavior by corporations.
method Analysis of a large US firm sample (2005-2019), focusing on selective disclosure.
result Religious adherence correlates with lower greenwashing behavior.
Machine learning method characterizes network interference in A/B tests.
problem Compromised A/B test reliability due to network interference.
method Causal network motifs and machine learning models.
result Outperforms conventional methods in characterizing network interference.
ScoreStop uses gradient tests to stop gradient boosting early.
problem Overfitting in gradient boosted decision trees.
method ScoreStop uses a functional score test based on gradients to stop boosting.
result ScoreStop is competitive with loss-based early stopping methods.
Platform combines RL and language models to study narrative influence on AI decisions.
problem Understanding how narrative elements shape AI decision-making.
method Dual-system architecture with reinforcement learning and language model integration.
result Initial experiments show narrative frameworks can influence AI decision-making.
A new algorithm infers causal networks from data using topological thresholds.
problem Inferring causal networks from data.
method Two methods for determining topological thresholds: one to leave no disconnected nodes, the other to find a causal large connected component.
result The novel algorithm is faster and more accurate than the PC algorithm.
We propose to explain the predictions of a deep neural network, by pointing to the set of what we call representer points in the training set, for a given test point prediction. Specifically, we show that we can decompose the pre-activation prediction of a neural network into a linear combination of activations of trai…
Generalizes causal inference to high-dimensional outcomes.
problem Limited causal inference methods for multivariate outcomes.
method Formulates causal discrepancy tests for nominal variables, uses conditional independence tests.
result Causal CDcorr method improves finite sample validity and power.