We present AutoSchemaKG, a framework for fully autonomous knowledge graph construction that eliminates the need for predefined schemas. Our system leverages large language models to simultaneously extract knowledge triples and induce comprehensive schemas directly from text, modeling both entities and events while employing conceptualization to organize instances into semantic categories. Processing over 50 million documents, we construct ATLAS (Automated Triple Linking And Schema induction), a family of knowledge graphs with 900+ million nodes and 5.9 billion edges. This approach outperforms state-of-the-art baselines on multi-hop QA tasks and enhances LLM factuality. Notably, our schema induction achieves 92% semantic alignment with human-crafted schemas with zero manual intervention, demonstrating that billion-scale knowledge graphs with dynamically induced schemas can effectively complement parametric knowledge in large language models.
Uncertainty quantification is crucial in safety-critical systems, where decisions must be made under uncertainty. In particular, we consider the problem of online uncertainty quantification, where data points arrive sequentially. Online conformal prediction is a principled online uncertainty quantification method that dynamically constructs a prediction set at each time step. While existing methods for online conformal prediction provide long-run coverage guarantees without any distributional assumptions, they typically assume a full feedback setting in which the true label is always observed. In this paper, we propose a novel learning method for online conformal prediction with partial feedback from an adaptive adversary-a more challenging setup where the true label is revealed only when it lies inside the constructed prediction set. Specifically, we formulate online conformal prediction as an adversarial bandit problem by treating each candidate prediction set as an arm. Building on an existing algorithm for adversarial bandits, our method achieves a long-run coverage guarantee by explicitly establishing its connection to the regret of the learner. Finally, we empirically demonstrate the effectiveness of our method in both independent and identically distributed (i.i.d.) and non-i.i.d. settings, showing that it successfully controls the miscoverage rate while maintaining a reasonable size of the prediction set.
How to efficiently perform network tomography is a fundamental problem in network management and monitoring. A network tomography task usually consists of applying multiple probing experiments, e.g., across different paths or via different casts (e.g., unicast and multicast). We study how to optimize the network tomography process through online sequential decisionmaking. From the methodology perspective, we introduce an online probe allocation algorithm that sequentially performs network tomography based on the principles of optimal experimental design and the maximum likelihood estimation. We rigorously analyze the regret of the algorithm under the conditions that $i)$ the optimal allocation is Lipschitz continuous in the parameters being estimated and ii) the parameter estimators satisfy a concentration property. From the application perspective, we demonstrate that the quantum bit-flip network fulfills the two theoretical conditions and provide their corresponding regrets when deploying our proposed online probe allocation algorithm. Besides case studies with theoretical guarantees, we also conduct simulations to compare our proposed algorithm with existing methods and demonstrate our algorithm's effectiveness.
Segmentation plays a crucial role in the automated morphological analysis of blood smear images, as it directly impacts the accuracy of various operations like cell count determination and disease detection. However, existing computerized methods for cell segmentation have not been able to achieve the same level of accuracy as expert histopathologists. This is due to the inherent challenges posed by the diverse shapes, sizes, and morphological characteristics of blood cells, the presence of clumped regions, and variations in slide preparation. To address this issue, we propose a double encoder-decoder network called "LeukoSegmenter" that focuses on segmenting leukocytes and diagnosing diseases like leukemia. The LeukoSegmenter network architecture consists of two encoder-decoder pairs (EDPs) of U-Nets cascaded in series. The cascading design refines pixel-level information, enhancing feature extraction while avoiding overfitting. Each EDP adopts a modified U-Net architecture with ResNet18 as the encoder, leveraging transfer learning to extract hierarchical features efficiently. Skip connections within the U-Net preserve fine details and improve gradient flow. The decoder upscales feature maps using transposed convolutions, restoring spatial resolution and generating a coarse segmentation map in the first EDP. This output is concatenated with the original image and passed to the second EDP. This approach enables the latter EDP to specifically concentrate on leukocytes while disregarding other blood cells and debris, thereby enhancing the segmentation accuracy. The proposed method achieves DICE scores of 95.19%, 94.20%, and 95.71%; IoU of 92.68%, 94.82%, and 94.24%; precision of 96.58%, 95.29%, and 94.45%; and accuracy of 95.77%, 94.22%, and 93.55% on ALL-IDB1, LISC, and Cellavision datasets, respectively. The model also demonstrated resilience to different types of noise and blurring effects.
Matrix mechanisms are often used to provide unbiased differentially private query answers when publishing statistics or creating synthetic data. Recent work has developed matrix mechanisms, such as ResidualPlanner and Weighted Fourier Factorizations, that scale to high dimensional datasets while providing optimality guarantees for workloads such as marginals and circular product queries. They operate by adding noise to a linearly independent set of queries that can compactly represent the desired workloads. In this paper, we present QuerySmasher, an alternative scalable approach based on a divide-and-conquer strategy. Given a workload that can be answered from various data marginals, QuerySmasher splits each query into sub-queries and re-assembles the pieces into mutually orthogonal sub-workloads. These sub-workloads represent small, low-dimensional problems that can be independently and optimally answered by existing low-dimensional matrix mechanisms. QuerySmasher then stitches these solutions together to answer queries in the original workload. We show that QuerySmasher subsumes prior work, like ResidualPlanner (RP), ResidualPlanner+ (RP+), and Weighted Fourier Factorizations (WFF). We prove that it can dominate those approaches, under sum squared error, for all workloads. We also experimentally demonstrate the scalability and accuracy of QuerySmasher.
