Language-supervised vision models have recently attracted great attention in computer vision. A common approach to build such models is to use contrastive learning on paired data across the two modalities, as exemplified by Contrastive Language-Image Pre-Training (CLIP). In this paper, under linear representation settings, (i) we initiate the investigation of a general class of nonlinear loss functions for multimodal contrastive learning (MMCL) including CLIP loss and show its connection to singular value decomposition (SVD). Namely, we show that each step of loss minimization by gradient descent can be seen as performing SVD on a contrastive cross-covariance matrix. Based on this insight, (ii) we analyze the performance of MMCL. We quantitatively show that the feature learning ability of MMCL can be better than that of unimodal contrastive learning applied to each modality even under the presence of wrongly matched pairs. This characterizes the robustness of MMCL to noisy data. Furthermore, when we have access to additional unpaired data, (iii) we propose a new MMCL loss that incorporates additional unpaired datasets. We show that the algorithm can detect the ground-truth pairs and improve performance by fully exploiting unpaired datasets. The performance of the proposed algorithm was verified by numerical experiments.

This paper studies offline policy learning, which aims at utilizing observations collected a priori (from either fixed or adaptively evolving behavior policies) to learn the optimal individualized decision rule in a given class. Existing policy learning methods rely on a uniform overlap assumption, i.e., the propensities of exploring all actions for all individual characteristics are lower bounded in the offline dataset. In other words, the performance of these methods depends on the worst-case propensity in the offline dataset. As one has no control over the data collection process, this assumption can be unrealistic in many situations, especially when the behavior policies are allowed to evolve over time with diminishing propensities. In this paper, we propose a new algorithm that optimizes lower confidence bounds (LCBs) -- instead of point estimates -- of the policy values. The LCBs are constructed by quantifying the estimation uncertainty of the augmented inverse propensity weighted (AIPW)-type estimators using knowledge of the behavior policies for collecting the offline data. Without assuming any uniform overlap condition, we establish a data-dependent upper bound for the suboptimality of our algorithm, which depends only on (i) the overlap for the optimal policy, and (ii) the complexity of the policy class. As an implication, for adaptively collected data, we ensure efficient policy learning as long as the propensities for optimal actions are lower bounded over time, while those for suboptimal ones are allowed to diminish arbitrarily fast. In our theoretical analysis, we develop a new self-normalized concentration inequality for IPW estimators, generalizing the well-known empirical Bernstein's inequality to unbounded and non-i.i.d. data.

Fractals are geometric shapes that can display complex and self-similar patterns found in nature (e.g., clouds and plants). Recent works in visual recognition have leveraged this property to create random fractal images for model pre-training. In this paper, we study the inverse problem -- given a target image (not necessarily a fractal), we aim to generate a fractal image that looks like it. We propose a novel approach that learns the parameters underlying a fractal image via gradient descent. We show that our approach can find fractal parameters of high visual quality and be compatible with different loss functions, opening up several potentials, e.g., learning fractals for downstream tasks, scientific understanding, etc.

We develop and analyze a principled approach to kernel ridge regression under covariate shift. The goal is to learn a regression function with small mean squared error over a target distribution, based on unlabeled data from there and labeled data that may have a different feature distribution. We propose to split the labeled data into two subsets and conduct kernel ridge regression on them separately to obtain a collection of candidate models and an imputation model. We use the latter to fill the missing labels and then select the best candidate model accordingly. Our non-asymptotic excess risk bounds show that in quite general scenarios, our estimator adapts to the structure of the target distribution as well as the covariate shift. It achieves the minimax optimal error rate up to a logarithmic factor. The use of pseudo-labels in model selection does not have major negative impacts.

I propose a novel dual-attention model(DAM) for aspect-level sentiment classification. Many methods have been proposed, such as support vector machines for artificial design features, long short-term memory networks based on attention mechanisms, and graph neural networks based on dependency parsing. While these methods all have decent performance, I think they all miss one important piece of syntactic information: dependency labels. Based on this idea, this paper proposes a model using dependency labels for the attention mechanism to do this task. We evaluate the proposed approach on three datasets: laptop and restaurant are from SemEval 2014, and the last one is a twitter dataset. Experimental results show that the dual attention model has good performance on all three datasets.

Multi-view clustering has been widely used in recent years in comparison to single-view clustering, for clear reasons, as it offers more insights into the data, which has brought with it some challenges, such as how to combine these views or features. Most of recent work in this field focuses mainly on tensor representation instead of treating the data as simple matrices. This permits to deal with the high-order correlation between the data which the based matrix approach struggles to capture. Accordingly, we will examine and compare these approaches, particularly in two categories, namely graph-based clustering and subspace-based clustering. We will conduct and report experiments of the main clustering methods over a benchmark datasets.

We propose a class of models based on Fisher's Linear Discriminant (FLD) for domain adaptation. The class entails a convex combination of two hypotheses: i) an average hypothesis representing previously encountered source tasks and ii) a hypothesis trained on a new target task. For a particular generative setting, we derive the expected risk of this combined hypothesis with respect to the target distribution and propose a computable approximation. This is then leveraged to estimate an optimal convex coefficient that exploits the bias-variance trade-off between source and target information to arrive at an optimal classifier for the target task. We study the effect of various generative parameter settings on the relative risks between the optimal hypothesis, hypothesis i), and hypothesis ii). Furthermore, we demonstrate the effectiveness of the proposed optimal classifier in several EEGand ECG-based classification problems and argue that the optimal classifier can be computed without access to direct information from any of the individual source tasks, leading to the preservation of privacy. We conclude by discussing further applications, limitations, and potential future directions. In problems with limited context-specific labeled data, machine learning models often fail to generalize well. These approaches are either ineffective or unavailable for problems where the input signals are highly variable across contexts or where a single model does not have access to a sufficient amount of data due to privacy or resource constraints (Mühlhoff, 2021). Note that the terms "context" and "task" can be used interchangeably here.

The performance of fault diagnosis systems is highly affected by data quality in cyber-physical power systems. These systems generate massive amounts of data that overburden the system with excessive computational costs. Another issue is the presence of noise in recorded measurements, which prevents building a precise decision model. Furthermore, the diagnostic model is often provided with a mixture of redundant measurements that may deviate it from learning normal and fault distributions. This paper presents the effect of feature engineering on mitigating the aforementioned challenges in cyber-physical systems. Feature selection and dimensionality reduction methods are combined with decision models to simulate data-driven fault diagnosis in a 118-bus power system. A comparative study is enabled accordingly to compare several advanced techniques in both domains. Dimensionality reduction and feature selection methods are compared both jointly and separately. Finally, experiments are concluded, and a setting is suggested that enhances data quality for fault diagnosis.

Online Class-Incremental (OCI) learning has sparked new approaches to expand the previously trained model knowledge from sequentially arriving data streams with new classes. Unfortunately, OCI learning can suffer from catastrophic forgetting (CF) as the decision boundaries for old classes can become inaccurate when perturbated by new ones. Existing literature have applied the data augmentation (DA) to alleviate the model forgetting, while the role of DA in OCI has not been well understood so far. In this paper, we theoretically show that augmented samples with lower correlation to the original data are more effective in preventing forgetting. However, aggressive augmentation may also reduce the consistency between data and corresponding labels, which motivates us to exploit proper DA to boost the OCI performance and prevent the CF problem. We propose the Enhanced Mixup (EnMix) method that mixes the augmented samples and their labels simultaneously, which is shown to enhance the sample diversity while maintaining strong consistency with corresponding labels. Further, to solve the class imbalance problem, we design an Adaptive Mixup (AdpMix) method to calibrate the decision boundaries by mixing samples from both old and new classes and dynamically adjusting the label mixing ratio. Our approach is demonstrated to be effective on several benchmark datasets through extensive experiments, and it is shown to be compatible with other replay-based techniques.

Model parameter regularization is a widely used technique to improve generalization, but also can be used to shape the weight distributions for various purposes. In this work, we shed light on how weight regularization can assist model quantization and compression techniques, and then propose range regularization (R^2) to further boost the quality of model optimization by focusing on the outlier prevention. By effectively regulating the minimum and maximum weight values from a distribution, we mold the overall distribution into a tight shape so that model compression and quantization techniques can better utilize their limited numeric representation powers. We introduce L-inf regularization, its extension margin regularization and a new soft-min-max regularization to be used as a regularization loss during full-precision model training. Coupled with state-of-the-art quantization and compression techniques, models trained with R^2 perform better on an average, specifically at lower bit weights with 16x compression ratio. We also demonstrate that R^2 helps parameter constrained models like MobileNetV1 achieve significant improvement of around 8% for 2 bit quantization and 7% for 1 bit compression.