Machine learning (ML) is playing an increasing role in decision-making tasks that directly affect individuals, e.g., loan approvals, or job applicant screening. Significant concerns arise that, without special provisions, individuals from under-privileged backgrounds may not get equitable access to services and opportunities. Existing research studies fairness with respect to protected attributes such as gender, race or income, but the impact of location data on fairness has been largely overlooked. With the widespread adoption of mobile apps, geospatial attributes are increasingly used in ML, and their potential to introduce unfair bias is significant, given their high correlation with protected attributes. We propose techniques to mitigate location bias in machine learning. Specifically, we consider the issue of miscalibration when dealing with geospatial attributes. We focus on spatial group fairness and we propose a spatial indexing algorithm that accounts for fairness. Our KD-tree inspired approach significantly improves fairness while maintaining high learning accuracy, as shown by extensive experimental results on real data.

The effectiveness of machine learning models is significantly affected by the size of the dataset and the quality of features as redundant and irrelevant features can radically degrade the performance. This paper proposes IGRF-RFE: a hybrid feature selection method tasked for multi-class network anomalies using a Multilayer perceptron (MLP) network. IGRF-RFE can be considered as a feature reduction technique based on both the filter feature selection method and the wrapper feature selection method. In our proposed method, we use the filter feature selection method, which is the combination of Information Gain and Random Forest Importance, to reduce the feature subset search space. Then, we apply recursive feature elimination(RFE) as a wrapper feature selection method to further eliminate redundant features recursively on the reduced feature subsets. Our experimental results obtained based on the UNSW-NB15 dataset confirm that our proposed method can improve the accuracy of anomaly detection while reducing the feature dimension. The results show that the feature dimension is reduced from 42 to 23 while the multi-classification accuracy of MLP is improved from 82.25% to 84.24%.

Cybersickness is a common ailment associated with virtual reality (VR) user experiences. Several automated methods exist based on machine learning (ML) and deep learning (DL) to detect cybersickness. However, most of these cybersickness detection methods are perceived as computationally intensive and black-box methods. Thus, those techniques are neither trustworthy nor practical for deploying on standalone energy-constrained VR head-mounted devices (HMDs). In this work, we present an explainable artificial intelligence (XAI)-based framework, LiteVR, for cybersickness detection, explaining the model's outcome and reducing the feature dimensions and overall computational costs. First, we develop three cybersickness DL models based on long-term short-term memory (LSTM), gated recurrent unit (GRU), and multilayer perceptron (MLP). Then, we employed a post-hoc explanation, such as SHapley Additive Explanations (SHAP), to explain the results and extract the most dominant features of cybersickness. Finally, we retrain the DL models with the reduced number of features. Our results show that eye-tracking features are the most dominant for cybersickness detection. Furthermore, based on the XAI-based feature ranking and dimensionality reduction, we significantly reduce the model's size by up to 4.3x, training time by up to 5.6x, and its inference time by up to 3.8x, with higher cybersickness detection accuracy and low regression error (i.e., on Fast Motion Scale (FMS)). Our proposed lite LSTM model obtained an accuracy of 94% in classifying cybersickness and regressing (i.e., FMS 1-10) with a Root Mean Square Error (RMSE) of 0.30, which outperforms the state-of-the-art. Our proposed LiteVR framework can help researchers and practitioners analyze, detect, and deploy their DL-based cybersickness detection models in standalone VR HMDs.

Spatially resolved transcriptomics brings exciting breakthroughs to single-cell analysis by providing physical locations along with gene expression. However, as a cost of the extremely high spatial resolution, the cellular level spatial transcriptomic data suffer significantly from missing values. While a standard solution is to perform imputation on the missing values, most existing methods either overlook spatial information or only incorporate localized spatial context without the ability to capture long-range spatial information. Using multi-head self-attention mechanisms and positional encoding, transformer models can readily grasp the relationship between tokens and encode location information. In this paper, by treating single cells as spatial tokens, we study how to leverage transformers to facilitate spatial tanscriptomics imputation. In particular, investigate the following two key questions: (1) $\textit{how to encode spatial information of cells in transformers}$, and (2) $\textit{ how to train a transformer for transcriptomic imputation}$. By answering these two questions, we present a transformer-based imputation framework, SpaFormer, for cellular-level spatial transcriptomic data. Extensive experiments demonstrate that SpaFormer outperforms existing state-of-the-art imputation algorithms on three large-scale datasets.

The construction of machine learning models involves many bi-level multi-objective optimization problems (BL-MOPs), where upper level (UL) candidate solutions must be evaluated via training weights of a model in the lower level (LL). Due to the Pareto optimality of sub-problems and the complex dependency across UL solutions and LL weights, an UL solution is feasible if and only if the LL weight is Pareto optimal. It is computationally expensive to determine which LL Pareto weight in the LL Pareto weight set is the most appropriate for each UL solution. This paper proposes a bi-level multi-objective learning framework (BLMOL), coupling the above decision-making process with the optimization process of the UL-MOP by introducing LL preference $r$. Specifically, the UL variable and $r$ are simultaneously searched to minimize multiple UL objectives by evolutionary multi-objective algorithms. The LL weight with respect to $r$ is trained to minimize multiple LL objectives via gradient-based preference multi-objective algorithms. In addition, the preference surrogate model is constructed to replace the expensive evaluation process of the UL-MOP. We consider a novel case study on multi-task graph neural topology search. It aims to find a set of Pareto topologies and their Pareto weights, representing different trade-offs across tasks at UL and LL, respectively. The found graph neural network is employed to solve multiple tasks simultaneously, including graph classification, node classification, and link prediction. Experimental results demonstrate that BLMOL can outperform some state-of-the-art algorithms and generate well-representative UL solutions and LL weights.

Bayesian learning via Stochastic Gradient Langevin Dynamics (SGLD) has been suggested for differentially private learning. While previous research provides differential privacy bounds for SGLD at the initial steps of the algorithm or when close to convergence, the question of what differential privacy guarantees can be made in between remains unanswered. This interim region is of great importance, especially for Bayesian neural networks, as it is hard to guarantee convergence to the posterior. This paper shows that using SGLD might result in unbounded privacy loss for this interim region, even when sampling from the posterior is as differentially private as desired.

The stochastic heavy ball method (SHB), also known as stochastic gradient descent (SGD) with Polyak's momentum, is widely used in training neural networks. However, despite the remarkable success of such algorithm in practice, its theoretical characterization remains limited. In this paper, we focus on neural networks with two and three layers and provide a rigorous understanding of the properties of the solutions found by SHB: \emph{(i)} stability after dropping out part of the neurons, \emph{(ii)} connectivity along a low-loss path, and \emph{(iii)} convergence to the global optimum. To achieve this goal, we take a mean-field view and relate the SHB dynamics to a certain partial differential equation in the limit of large network widths. This mean-field perspective has inspired a recent line of work focusing on SGD while, in contrast, our paper considers an algorithm with momentum. More specifically, after proving existence and uniqueness of the limit differential equations, we show convergence to the global optimum and give a quantitative bound between the mean-field limit and the SHB dynamics of a finite-width network. Armed with this last bound, we are able to establish the dropout-stability and connectivity of SHB solutions.

Due to their conceptual simplicity, k-means algorithm variants have been extensively used for unsupervised cluster analysis. However, one main shortcoming of these algorithms is that they essentially fit a mixture of identical spherical Gaussians to data that vastly deviates from such a distribution. In comparison, general Gaussian Mixture Models (GMMs) can fit richer structures but require estimating a quadratic number of parameters per cluster to represent the covariance matrices. This poses two main issues: (i) the underlying optimization problems are challenging due to their larger number of local minima, and (ii) their solutions can overfit the data. In this work, we design search strategies that circumvent both issues. We develop efficient global optimization algorithms for general GMMs, and we combine these algorithms with regularization strategies that avoid overfitting. Through extensive computational analyses, we observe that global optimization or regularization in isolation does not substantially improve cluster recovery. However, combining these techniques permits a completely new level of performance previously unachieved by k-means algorithm variants, unraveling vastly different cluster structures. These results shed new light on the current status quo between GMM and k-means methods and suggest the more frequent use of general GMMs for data exploration. To facilitate such applications, we provide open-source code as well as Julia packages ("UnsupervisedClustering.jl" and "RegularizedCovarianceMatrices.jl") implementing the proposed techniques.

Dealing with distribution shifts is one of the central challenges for modern machine learning. One fundamental situation is the \emph{covariate shift}, where the input distributions of data change from training to testing stages while the input-conditional output distribution remains unchanged. In this paper, we initiate the study of a more challenging scenario -- \emph{continuous} covariate shift -- in which the test data appear sequentially, and their distributions can shift continuously. Our goal is to adaptively train the predictor such that its prediction risk accumulated over time can be minimized. Starting with the importance-weighted learning, we show the method works effectively if the time-varying density ratios of test and train inputs can be accurately estimated. However, existing density ratio estimation methods would fail due to data scarcity at each time step. To this end, we propose an online method that can appropriately reuse historical information. Our density ratio estimation method is proven to perform well by enjoying a dynamic regret bound, which finally leads to an excess risk guarantee for the predictor. Empirical results also validate the effectiveness.

Prior work has successfully incorporated optimization layers as the last layer in neural networks for various problems, thereby allowing joint learning and planning in one neural network forward pass. In this work, we identify a weakness in such a set-up where inputs to the optimization layer lead to undefined output of the neural network. Such undefined decision outputs can lead to possible catastrophic outcomes in critical real time applications. We show that an adversary can cause such failures by forcing rank deficiency on the matrix fed to the optimization layer which results in the optimization failing to produce a solution. We provide a defense for the failure cases by controlling the condition number of the input matrix. We study the problem in the settings of synthetic data, Jigsaw Sudoku, and in speed planning for autonomous driving, building on top of prior frameworks in end-to-end learning and optimization. We show that our proposed defense effectively prevents the framework from failing with undefined output. Finally, we surface a number of edge cases which lead to serious bugs in popular equation and optimization solvers which can be abused as well.