Machine Learning (stat.ML)

Abstract: The mean shift (MS) algorithm seeks a mode of the kernel density estimate (KDE). This study presents a convergence guarantee of the mode estimate sequence generated by the MS algorithm and an evaluation of the convergence rate, under fairly mild conditions, with the help of the argument concerning the {\L}ojasiewicz inequality. Our findings, which extend existing ones covering analytic kernels and the Epanechnikov kernel, are significant in that they cover the biweight kernel that is optimal among non-negative kernels in terms of the asymptotic statistical efficiency for the KDE-based mode estimation.

Abstract: Label smoothing (LS) adopts smoothed targets in classification tasks. For example, in binary classification, instead of the one-hot target $(1,0)^\top$ used in conventional logistic regression (LR), LR with LS (LSLR) uses the smoothed target $(1-\frac{\alpha}{2},\frac{\alpha}{2})^\top$ with a smoothing level $\alpha\in(0,1)$, which causes squeezing of values of the logit. Apart from the common regularization-based interpretation of LS that leads to an inconsistent probability estimator, we regard LSLR as modifying the loss function and consistent estimator for probability estimation. In order to study the significance of each of these two modifications by LSLR, we introduce a modified LSLR (MLSLR) that uses the same loss function as LSLR and the same consistent estimator as LR, while not squeezing the logits. For the loss function modification, we theoretically show that MLSLR with a larger smoothing level has lower efficiency with correctly-specified models, while it exhibits higher robustness against model misspecification than LR. Also, for the modification of the probability estimator, an experimental comparison between LSLR and MLSLR showed that this modification and squeezing of the logits in LSLR have negative effects on the probability estimation and classification performance. The understanding of the properties of LS provided by these comparisons allows us to propose MLSLR as an improvement over LSLR.

Abstract: Supervised learning is often affected by a covariate shift in which the marginal distributions of instances (covariates $x$) of training and testing samples $\mathrm{p}_\text{tr}(x)$ and $\mathrm{p}_\text{te}(x)$ are different but the label conditionals coincide. Existing approaches address such covariate shift by either using the ratio $\mathrm{p}_\text{te}(x)/\mathrm{p}_\text{tr}(x)$ to weight training samples (reweighting methods) or using the ratio $\mathrm{p}_\text{tr}(x)/\mathrm{p}_\text{te}(x)$ to weight testing samples (robust methods). However, the performance of such approaches can be poor under support mismatch or when the above ratios take large values. We propose a minimax risk classification (MRC) approach for covariate shift adaptation that avoids such limitations by weighting both training and testing samples. In addition, we develop effective techniques that obtain both sets of weights and generalize the conventional kernel mean matching method. We provide novel generalization bounds for our method that show a significant increase in the effective sample size compared with reweighted methods. The proposed method also achieves enhanced classification performance in both synthetic and empirical experiments.

Abstract: With the increasing adoption of AI-based systems across everyday life, the need to understand their decision-making mechanisms is correspondingly accelerating. The level at which we can trust the statistical inferences made from AI-based decision systems is an increasing concern, especially in high-risk systems such as criminal justice or medical diagnosis, where incorrect inferences may have tragic consequences. Despite their successes in providing solutions to problems involving real-world data, deep learning (DL) models cannot quantify the certainty of their predictions. And are frequently quite confident, even when their solutions are incorrect. This work presents a method to infer prominent features in two DL classification models trained on clinical and non-clinical text by employing techniques from topological and geometric data analysis. We create a graph of a model's prediction space and cluster the inputs into the graph's vertices by the similarity of features and prediction statistics. We then extract subgraphs demonstrating high-predictive accuracy for a given label. These subgraphs contain a wealth of information about features that the DL model has recognized as relevant to its decisions. We infer these features for a given label using a distance metric between probability measures, and demonstrate the stability of our method compared to the LIME interpretability method. This work demonstrates that we may gain insights into the decision mechanism of a DL model, which allows us to ascertain if the model is making its decisions based on information germane to the problem or identifies extraneous patterns within the data.

Abstract: Data from populations of systems are prevalent in many industrial applications. Machines and infrastructure are increasingly instrumented with sensing systems, emitting streams of telemetry data with complex interdependencies. In practice, data-centric monitoring procedures tend to consider these assets (and respective models) as distinct -- operating in isolation and associated with independent data. In contrast, this work captures the statistical correlations and interdependencies between models of a group of systems. Utilising a Bayesian multilevel approach, the value of data can be extended, since the population can be considered as a whole, rather than constituent parts. Most interestingly, domain expertise and knowledge of the underlying physics can be encoded in the model at the system, subgroup, or population level. We present an example of acoustic emission (time-of-arrival) mapping for source location, to illustrate how multilevel models naturally lend themselves to representing aggregate systems in engineering. In particular, we focus on constraining the combined models with domain knowledge to enhance transfer learning and enable further insights at the population level.

Abstract: Information spread through social networks is ubiquitous. Influence maximiza- tion (IM) algorithms aim to identify individuals who will generate the greatest spread through the social network if provided with information, and have been largely devel- oped with marketing in mind. In social networks with community structure, which are very common, IM algorithms focused solely on maximizing spread may yield signifi- cant disparities in information coverage between communities, which is problematic in settings such as public health messaging. While some IM algorithms aim to remedy disparity in information coverage using node attributes, none use the empirical com- munity structure within the network itself, which may be beneficial since communities directly affect the spread of information. Further, the use of empirical network struc- ture allows us to leverage community detection techniques, making it possible to run fair-aware algorithms when there are no relevant node attributes available, or when node attributes do not accurately capture network community structure. In contrast to other fair IM algorithms, this work relies on fitting a model to the social network which is then used to determine a seed allocation strategy for optimal fair information spread. We develop an algorithm to determine optimal seed allocations for expected fair coverage, defined through maximum entropy, provide some theoretical guarantees under appropriate conditions, and demonstrate its empirical accuracy on both simu- lated and real networks. Because this algorithm relies on a fitted network model and not on the network directly, it is well-suited for partially observed and noisy social networks.