arXiv subjects

David Zhao

Publications and source records attributed to David Zhao.

15 recordsLinked to original sources

arXiv2025

Rel: A Programming Language for Relational Data

From the moment of their inception, languages for relational data have been described as sublanguages embedded in a host programming language. Rel is a new relational language whose key design goal is to go beyond this paradigm with features that allow for programming in the large, making it possible to fully describe end to end application semantics. With the new approach we can model the semantics of entire enterprise applications relationally, which helps significantly reduce architecture complexity and avoid the well-known impedance mismatch problem. This paradigm shift is enabled by 50 years of database research, making it possible to revisit the sublanguage/host language paradigm, starting from the fundamental principles. We present the main features of Rel: those that give it the power to express traditional query language operations and those that are designed to grow the language and allow programming in the large.

cs.DB↗

arXiv2025

Provenance Guided Rollback Suggestions

Advances in incremental Datalog evaluation strategies have made Datalog popular among use cases with constantly evolving inputs such as static analysis in continuous integration and deployment pipelines. As a result, new logic programming debugging techniques are needed to support these emerging use cases. This paper introduces an incremental debugging technique for Datalog, which determines the failing changes for a \emph{rollback} in an incremental setup. Our debugging technique leverages a novel incremental provenance method. We have implemented our technique using an incremental version of the Souffl\'{e} Datalog engine and evaluated its effectiveness on the DaCapo Java program benchmarks analyzed by the Doop static analysis library. Compared to state-of-the-art techniques, we can localize faults and suggest rollbacks with an overall speedup of over 26.9$\times$ while providing higher quality results.

cs.LO↗

arXiv2022

Multiscale modeling of solute diffusion in triblock copolymer membranes

We develop a multiscale simulation model for diffusion of solutes through porous triblock copolymer membranes. The approach combines two techniques: self-consistent field theory (SCFT) to predict the structure of the self-assembled, solvated membrane and on-lattice kinetic Monte Carlo (kMC) simulations to model diffusion of solutes. Solvation is simulated in SCFT by constraining the glassy membrane matrix while relaxing the brush-like membrane pore coating against the solvent. The kMC simulations capture the resulting solute spatial distribution and concentration-dependent local diffusivity in the polymer-coated pores; we parameterize the latter using particle-based simulations. We apply our approach to simulate solute diffusion through nonequilibrium morphologies of a model triblock copolymer, and we correlate diffusivity with structural descriptors of the morphologies. We also compare the model's predictions to alternative approaches based on simple lattice random walks and find our multiscale model to be more robust and systematic to parameterize. Our multiscale modeling approach is general and can be readily extended in the future to other chemistries, morphologies, and models for the local solute diffusivity and interactions with the membrane.

cond-mat.soft↗

arXiv2022

Dynamic Global Sensitivity for Differentially Private Contextual Bandits

Bandit algorithms have become a reference solution for interactive recommendation. However, as such algorithms directly interact with users for improved recommendations, serious privacy concerns have been raised regarding its practical use. In this work, we propose a differentially private linear contextual bandit algorithm, via a tree-based mechanism to add Laplace or Gaussian noise to model parameters. Our key insight is that as the model converges during online update, the global sensitivity of its parameters shrinks over time (thus named dynamic global sensitivity). Compared with existing solutions, our dynamic global sensitivity analysis allows us to inject less noise to obtain $(\epsilon, \delta)$-differential privacy with added regret caused by noise injection in $\tilde O(\log{T}\sqrt{T}/\epsilon)$. We provide a rigorous theoretical analysis over the amount of noise added via dynamic global sensitivity and the corresponding upper regret bound of our proposed algorithm. Experimental results on both synthetic and real-world datasets confirmed the algorithm's advantage against existing solutions.

cs.LG↗

arXiv2022

Towards Instance-Wise Calibration: Local Amortized Diagnostics and Reshaping of Conditional Densities (LADaR)

Key science questions, such as galaxy distance estimation and weather forecasting, often require knowing the full predictive distribution of a target variable $y$ given complex inputs $\mathbf{x}$. Despite recent advances in machine learning and physics-based models, it remains challenging to assess whether an initial model is calibrated for all $\mathbf{x}$, and when needed, to reshape the densities of $y$ toward "instance-wise" calibration. This paper introduces the LADaR (Local Amortized Diagnostics and Reshaping of Conditional Densities) framework and proposes a new computationally efficient algorithm ($\texttt{Cal-PIT}$) that produces interpretable local diagnostics and provides a mechanism for adjusting conditional density estimates (CDEs). $\texttt{Cal-PIT}$ learns a single interpretable local probability--probability map from calibration data that identifies where and how the initial model is miscalibrated across feature space, which can be used to morph CDEs such that they are well-calibrated. We illustrate the LADaR framework on synthetic examples, including probabilistic forecasting from image sequences, akin to predicting storm wind speed from satellite imagery. Our main science application involves estimating the probability density functions of galaxy distances given photometric data, where $\texttt{Cal-PIT}$ achieves better instance-wise calibration than all 11 other literature methods in a benchmark data challenge, demonstrating its utility for next-generation cosmological analyses.

stat.ML↗

arXiv2021

Re-calibrating Photometric Redshift Probability Distributions Using Feature-space Regression

Many astrophysical analyses depend on estimates of redshifts (a proxy for distance) determined from photometric (i.e., imaging) data alone. Inaccurate estimates of photometric redshift uncertainties can result in large systematic errors. However, probability distribution outputs from many photometric redshift methods do not follow the frequentist definition of a Probability Density Function (PDF) for redshift -- i.e., the fraction of times the true redshift falls between two limits $z_{1}$ and $z_{2}$ should be equal to the integral of the PDF between these limits. Previous works have used the global distribution of Probability Integral Transform (PIT) values to re-calibrate PDFs, but offsetting inaccuracies in different regions of feature space can conspire to limit the efficacy of the method. We leverage a recently developed regression technique that characterizes the local PIT distribution at any location in feature space to perform a local re-calibration of photometric redshift PDFs. Though we focus on an example from astrophysics, our method can produce PDFs which are calibrated at all locations in feature space for any use case.

astro-ph.IM↗

arXiv2021

The Choice Construct in the Souffle Language

Datalog has become a popular implementation language for solving large-scale, real-world problems, including bug finders, network analysis tools, and disassemblers. These applications express complex behaviour with hundreds of relations and rules that often require a non-deterministic choice for tuples in relations to express worklist algorithms. This work is an experience report that describes the implementation of a choice construct in the Datalog engine Souffle. With the choice construct, we can express worklist algorithms such as spanning trees in a few lines of code. We highlight the differences between rule-based choice as described in prior work, and relation-based choice introduced by this work. We show that a choice construct enables certain worklist algorithms to be computed up to 10kx faster than having no choice construct.

cs.PL↗

arXiv2021

Likelihood-Free Frequentist Inference: Bridging Classical Statistics and Machine Learning for Reliable Simulator-Based Inference

Many areas of science rely on simulators that implicitly encode intractable likelihood functions of complex systems. Classical statistical methods are poorly suited for these so-called likelihood-free inference (LFI) settings, especially outside asymptotic and low-dimensional regimes. At the same time, popular LFI methods - such as Approximate Bayesian Computation or more recent machine learning techniques - do not necessarily lead to valid scientific inference because they do not guarantee confidence sets with nominal coverage in general settings. In addition, LFI currently lacks practical diagnostic tools to check the actual coverage of computed confidence sets across the entire parameter space. In this work, we propose a modular inference framework that bridges classical statistics and modern machine learning to provide (i) a practical approach for constructing confidence sets with near finite-sample validity at any value of the unknown parameters, and (ii) interpretable diagnostics for estimating empirical coverage across the entire parameter space. We refer to this framework as likelihood-free frequentist inference (LF2I). Any method that defines a test statistic can leverage LF2I to create valid confidence sets and diagnostics without costly Monte Carlo or bootstrap samples at fixed parameter settings. We study two likelihood-based test statistics (ACORE and BFF) and demonstrate their performance on high-dimensional complex data. Code is available at https://github.com/lee-group-cmu/lf2i.

stat.ML↗

arXiv2021

MD-split+: Practical Local Conformal Inference in High Dimensions

Quantifying uncertainty in model predictions is a common goal for practitioners seeking more than just point predictions. One tool for uncertainty quantification that requires minimal assumptions is conformal inference, which can help create probabilistically valid prediction regions for black box models. Classical conformal prediction only provides marginal validity, whereas in many situations locally valid prediction regions are desirable. Deciding how best to partition the feature space X when applying localized conformal prediction is still an open question. We present MD-split+, a practical local conformal approach that creates X partitions based on localized model performance of conditional density estimation models. Our method handles complex real-world data settings where such models may be misspecified, and scales to high-dimensional inputs. We discuss how our local partitions philosophically align with expected behavior from an unattainable conditional conformal inference approach. We also empirically compare our method against other local conformal approaches.

stat.ML↗

arXiv2021

Diagnostics for Conditional Density Models and Bayesian Inference Algorithms

There has been growing interest in the AI community for precise uncertainty quantification. Conditional density models f(y|x), where x represents potentially high-dimensional features, are an integral part of uncertainty quantification in prediction and Bayesian inference. However, it is challenging to assess conditional density estimates and gain insight into modes of failure. While existing diagnostic tools can determine whether an approximated conditional density is compatible overall with a data sample, they lack a principled framework for identifying, locating, and interpreting the nature of statistically significant discrepancies over the entire feature space. In this paper, we present rigorous and easy-to-interpret diagnostics such as (i) the "Local Coverage Test" (LCT), which distinguishes an arbitrarily misspecified model from the true conditional density of the sample, and (ii) "Amortized Local P-P plots" (ALP) which can quickly provide interpretable graphical summaries of distributional differences at any location x in the feature space. Our validation procedures scale to high dimensions and can potentially adapt to any type of data at hand. We demonstrate the effectiveness of LCT and ALP through a simulated experiment and applications to prediction and parameter inference for image data.

stat.ME↗

arXiv2019

Reinforcement learning for bandwidth estimation and congestion control in real-time communications

Bandwidth estimation and congestion control for real-time communications (i.e., audio and video conferencing) remains a difficult problem, despite many years of research. Achieving high quality of experience (QoE) for end users requires continual updates due to changing network architectures and technologies. In this paper, we apply reinforcement learning for the first time to the problem of real-time communications (RTC), where we seek to optimize user-perceived quality. We present initial proof-of-concept results, where we learn an agent to control sending rate in an RTC system, evaluating using both network simulation and real Internet video calls. We discuss the challenges we observed, particularly in designing realistic reward functions that reflect QoE, and in bridging the gap between the training environment and real-world networks.

cs.NI↗

arXiv2019

Cryptocurrency Price Prediction and Trading Strategies Using Support Vector Machines

Few assets in financial history have been as notoriously volatile as cryptocurrencies. While the long term outlook for this asset class remains unclear, we are successful in making short term price predictions for several major crypto assets. Using historical data from July 2015 to November 2019, we develop a large number of technical indicators to capture patterns in the cryptocurrency market. We then test various classification methods to forecast short-term future price movements based on these indicators. On both PPV and NPV metrics, our classifiers do well in identifying up and down market moves over the next 1 hour. Beyond evaluating classification accuracy, we also develop a strategy for translating 1-hour-ahead class predictions into trading decisions, along with a backtester that simulates trading in a realistic environment. We find that support vector machines yield the most profitable trading strategies, which outperform the market on average for Bitcoin, Ethereum and Litecoin over the past 22 months, since January 2018.

q-fin.TR↗

arXiv2019

Provenance for Large-scale Datalog

Logic programming languages such as Datalog have become popular as Domain Specific Languages (DSLs) for solving large-scale, real-world problems, in particular, static program analysis and network analysis. The logic specifications which model analysis problems, process millions of tuples of data and contain hundreds of highly recursive rules. As a result, they are notoriously difficult to debug. While the database community has proposed several data-provenance techniques that address the Declarative Debugging Challenge for Databases, in the cases of analysis problems, these state-of-the-art techniques do not scale. In this paper, we introduce a novel bottom-up Datalog evaluation strategy for debugging: our provenance evaluation strategy relies on a new provenance lattice that includes proof annotations, and a new fixed-point semantics for semi-naive evaluation. A debugging query mechanism allows arbitrary provenance queries, constructing partial proof trees of tuples with minimal height. We integrate our technique into Souffle, a Datalog engine that synthesizes C++ code, and achieve high performance by using specialized parallel data structures. Experiments are conducted with DOOP/DaCapo, producing proof annotations for tens of millions of output tuples. We show that our method has a runtime overhead of 1.27x on average while being more flexible than existing state-of-the-art techniques.

cs.PL↗

arXiv2018

SAFE-PDF: Robust Detection of JavaScript PDF Malware Using Abstract Interpretation

The popularity of the PDF format and the rich JavaScript environment that PDF viewers offer make PDF documents an attractive attack vector for malware developers. PDF documents present a serious threat to the security of organizations because most users are unsuspecting of them and thus likely to open documents from untrusted sources. We propose to identify malicious PDFs by using conservative abstract interpretation to statically reason about the behavior of the embedded JavaScript code. Currently, state-of-the-art tools either: (1) statically identify PDF malware based on structural similarity to known malicious samples; or (2) dynamically execute the code to detect malicious behavior. These two approaches are subject to evasion attacks that mimic the structure of benign documents or do not exhibit their malicious behavior when being analyzed dynamically. In contrast, abstract interpretation is oblivious to both types of evasions. A comparison with two state-of-the-art PDF malware detection tools shows that our conservative abstract interpretation approach achieves similar accuracy, while being more resilient to evasion attacks.

cs.CR↗

arXiv2005

Coloring graphs with crossings

We generalize the Five Color Theorem by showing that it extends to graphs with two crossings. Furthermore, we show that if a graph has three crossings, but does not contain K_6 as a subgraph, then it is also 5-colorable. We also consider the question of whether the result can be extended to graphs with more crossings.

math.CO↗