arXiv subjects

Mihail Stoian

Publications and source records attributed to Mihail Stoian.

26 records · Page 2Linked to original sources

arXiv2024

Corra: Correlation-Aware Column Compression

Column encoding schemes have witnessed a spark of interest with the rise of open storage formats (like Parquet) in data lakes in modern cloud deployments. This is not surprising -- as data volume increases, it becomes more and more important to reduce storage cost on block storage (such as S3) as well as reduce memory pressure in multi-tenant in-memory buffers of cloud databases. However, single-column encoding schemes have reached a plateau in terms of the compression size they can achieve. We argue that this is due to the neglect of cross-column correlations. For instance, consider the column pair ($\texttt{city}$, $\texttt{zip_code}$). Typically, cities have only a few dozen unique zip codes. If this information is properly exploited, it can significantly reduce the space consumption of the latter column. In this work, we depart from the established path of compressing data using only single-column encoding schemes and introduce several what we call $\textit{horizontal}$, correlation-aware encoding schemes. We demonstrate their advantages over single-column encoding schemes on the well-known TPC-H's $\texttt{lineitem}$, LDBC's $\texttt{message}$, DMV, and Taxi datasets. Our correlation-aware encoding schemes save up to 58.3% of the compressed size over single-column schemes for $\texttt{lineitem}$'s $\texttt{receiptdate}$, 53.7% for DMV's $\texttt{zip_code}$, and 85.16% for Taxi's $\texttt{total_amount}$.

cs.DB↗

arXiv2024

TSP Escapes the $O(2^n n^2)$ Curse

The dynamic programming solution to the traveling salesman problem due to Bellman, and independently Held and Karp, runs in time $O(2^n n^2)$, with no improvement in the last sixty years. We break this barrier for the first time by designing an algorithm that runs in deterministic time $2^n n^2 / 2^{Ω(\sqrt{\log n})}$. We achieve this by strategically remodeling the dynamic programming recursion as a min-plus matrix product, for which faster-than-naïve algorithms exist.

cs.DS↗

arXiv2024

Did Fourier Really Meet Möbius? Fast Subset Convolution via FFT

In their seminal work on subset convolution, Björklund, Husfeldt, Kaski and Koivisto introduced the now well-known $O(2^n n^2)$-time evaluation of the subset convolution in the sum-product ring. This sparked a wave of remarkable results for fundamental problems, such as the minimum Steiner tree and the chromatic number. However, in spite of its theoretical improvement, large intermediate outputs and floating-point precision errors due to alternating addition and subtraction in its set function transforms make the algorithm unusable in practice. We provide a simple FFT-based algorithm that completely eliminates the need for set function transforms and maintains the running time of the original algorithm. This makes it possible to take advantage of nearly sixty years of research on efficient FFT implementations.

cs.DS↗

arXiv2024

The Art of Staying Ahead of Deadlines: Improved Algorithms for the Minimum Tardy Processing Time

We study the fundamental scheduling problem $1\|\sum p_jU_j$. Given a set of $n$ jobs with processing times $p_j$ and deadlines $d_j$, the problem is to select a subset of jobs such that the total processing time is maximized without violating the deadlines. In the midst of a flourishing line of research, Fischer and Wennmann have recently devised the sought-after $\widetilde O(P)$-time algorithm, where $P = \sum p_j$ is the total processing time of all jobs. This running time is optimal as it matches conditional lower bounds based on popular conjectures. However, $P$ is not the sole parameter one could parameterize the running time by. Indeed, they explicitly leave open the question of whether a running time of $\widetilde O(n + \max d_j)$ or even $\widetilde O(n + \max p_j)$ is possible. In this work, we show, somewhat surprisingly, that by a refined implementation of their original algorithm, one can obtain the asked-for $\widetilde O(n + \max d_j)$-time algorithm.

cs.DS↗

arXiv2021

Towards Practical Learned Indexing

Latest research proposes to replace existing index structures with learned models. However, current learned indexes tend to have many hyperparameters, often do not provide any error guarantees, and are expensive to build. We introduce Practical Learned Index (PLEX). PLEX only has a single hyperparameter $ε$ (maximum prediction error) and offers a better trade-off between build and lookup time than state-of-the-art approaches. Similar to RadixSpline, PLEX consists of a spline and a (multi-level) radix layer. It first builds a spline satisfying the given $ε$ and then performs an ad-hoc analysis of the distribution of spline points to quickly tune the radix layer.

cs.DB↗

arXiv2020

Benchmarking Learned Indexes

Recent advancements in learned index structures propose replacing existing index structures, like B-Trees, with approximate learned models. In this work, we present a unified benchmark that compares well-tuned implementations of three learned index structures against several state-of-the-art "traditional" baselines. Using four real-world datasets, we demonstrate that learned index structures can indeed outperform non-learned indexes in read-only in-memory workloads over a dense array. We also investigate the impact of caching, pipelining, dataset size, and key size. We study the performance profile of learned index structures, and build an explanation for why learned models achieve such good performance. Finally, we investigate other important properties of learned index structures, such as their performance in multi-threaded systems and their build times.

cs.DB↗

arXiv2020

RadixSpline: A Single-Pass Learned Index

Recent research has shown that learned models can outperform state-of-the-art index structures in size and lookup performance. While this is a very promising result, existing learned structures are often cumbersome to implement and are slow to build. In fact, most approaches that we are aware of require multiple training passes over the data. We introduce RadixSpline (RS), a learned index that can be built in a single pass over the data and is competitive with state-of-the-art learned index models, like RMI, in size and lookup performance. We evaluate RS using the SOSD benchmark and show that it achieves competitive results on all datasets, despite the fact that it only has two parameters.

cs.DB↗

arXiv2019

SOSD: A Benchmark for Learned Indexes

A groundswell of recent work has focused on improving data management systems with learned components. Specifically, work on learned index structures has proposed replacing traditional index structures, such as B-trees, with learned models. Given the decades of research committed to improving index structures, there is significant skepticism about whether learned indexes actually outperform state-of-the-art implementations of traditional structures on real-world data. To answer this question, we propose a new benchmarking framework that comes with a variety of real-world datasets and baseline implementations to compare against. We also show preliminary results for selected index structures, and find that learned models indeed often outperform state-of-the-art implementations, and are therefore a promising direction for future research.

cs.DB↗