In this paper, we introduce DobLIX, a dual-objective learned index (LI) specifically designed for Log-Structured Merge (LSM) tree-based key-value stores. Traditional LIs primarily focus on optimizing index lookups, often overlooking the critical role of data access from storage, which can become a significant performance bottleneck. In LSM-based systems, a considerable portion of the index is stored on disk, making lookups highly dependent on the efficient coordination between in-memory structures and disk-resident data. Poorly optimized access patterns can lead to excessive I/O operations, negatively impacting read latency and overall system performance. DobLIX addresses this by incorporating a second objective, data access optimization, into the LI training process. This dual-objective approach ensures that both index lookup efficiency and data access costs are minimized, leading to significant improvements in read performance while maintaining write efficiency in real-world LSM systems. Additionally, DobLIX features a reinforcement learning agent that dynamically tunes the system parameters, allowing it to adapt to varying workloads in real-time. Experimental results using real-world datasets demonstrate that DobLIX reduces indexing overhead and improves throughput by 1.19x to 2.21x compared to state-of-the-art methods within RocksDB, a widely used LSM-based storage engine.

We propose a novel framework for incorporating unlabeled data into semi-supervised classification problems, where scenarios involving the minimization of either i) adversarially robust or ii) non-robust loss functions have been considered. Notably, we allow the unlabeled samples to deviate slightly (in total variation sense) from the in-domain distribution. The core idea behind our framework is to combine Distributionally Robust Optimization (DRO) with self-supervised training. As a result, we also leverage efficient polynomial-time algorithms for the training stage. From a theoretical standpoint, we apply our framework on the classification problem of a mixture of two Gaussians in $\mathbb{R}^d$, where in addition to the $m$ independent and labeled samples from the true distribution, a set of $n$ (usually with $n\gg m$) out of domain and unlabeled samples are gievn as well. Using only the labeled data, it is known that the generalization error can be bounded by $\propto\left(d/m\right)^{1/2}$. However, using our method on both isotropic and non-isotropic Gaussian mixture models, one can derive a new set of analytically explicit and non-asymptotic bounds which show substantial improvement on the generalization error compared ERM. Our results underscore two significant insights: 1) out-of-domain samples, even when unlabeled, can be harnessed to narrow the generalization gap, provided that the true data distribution adheres to a form of the "cluster assumption", and 2) the semi-supervised learning paradigm can be regarded as a special case of our framework when there are no distributional shifts. We validate our claims through experiments conducted on a variety of synthetic and real-world datasets.

We introduce a few-shot learning framework for error detection. We show that data augmentation (a form of weak supervision) is key to training high-quality, ML-based error detection models that require minimal human involvement. Our framework consists of two parts: (1) an expressive model to learn rich representations that capture the inherent syntactic and semantic heterogeneity of errors; and (2) a data augmentation model that, given a small seed of clean records, uses dataset-specific transformations to automatically generate additional training data. Our key insight is to learn data augmentation policies from the noisy input dataset in a weakly supervised manner. We show that our framework detects errors with an average precision of ~94% and an average recall of ~93% across a diverse array of datasets that exhibit different types and amounts of errors. We compare our approach to a comprehensive collection of error detection methods, ranging from traditional rule-based methods to ensemble-based and active learning approaches. We show that data augmentation yields an average improvement of 20 F1 points while it requires access to 3x fewer labeled examples compared to other ML approaches.

An optimal placement of protective devices could increase the reliability and quality level of a distribution network. An innovative mixed-integer nonlinear programming model is proposed in this paper to find the type, optimal siting, and number of protective devices to be accurately installed in distribution networks. The customer outage and protective devices costs are considered to derive a value-based reliability equation. To ensure the effectiveness of the proposed formulation economic and technical constraints is considered. Further, this paper aims at aiding decision-makers in providing appropriate protective device allocation by minimizing the expected interruption cost index. Case studies are employed to demonstrate the reliability optimization of a test network and a typical real-size network in which the several cost constraints and protection schemes are assumed to extract the results. Accuracy and effectiveness of the proposed method are assessed and sensitivities analysis is carried out.