BZPEERReading, trust and collaboration for research
Menu

Discover

GraphFollow connections around a paper, topic or saved view.

Workspaces

HomePick up where your research flow left off.InboxHandle requests, replies and notifications from one place.CollectionsCurate reading lists, evidence sets and shareable dossiers.ResearchSee your private provenance graph, trail and imports.CollaborationMove from discovery into focused discussion rooms.PublishDraft, preview and publish full-text research articles.

Network

ExpertsFind specialists worth following or asking for help.CommunitiesJoin research circles organized around shared work.PartnersSee external organizations active around the same topics.

Opportunities

ListingsBrowse roles, calls and collaborations with clearer fit signals.

Account

Log inReturn to your reading and collaboration workspace.Create accountStart saving work, following people and joining teams.
Log inCreate account

Researcher profile

Marco Molinaro

Marco Molinaro contributes to research discovery and scholarly infrastructure.

ResearcherAffiliation not importedOpen to collaborate
math.OCData Structures and AlgorithmsMachine Learningastro-ph.SRastro-ph.EPastro-ph.IMArtificial IntelligenceComputational Complexity

Trust snapshot

Quick read

Trust 21 - EmergingVerification L1Unclaimed author
15works
0followers
8topics
4close collaborators

Actions

Decide how to stay connected

Follow researcher0

Identity and collaboration

How to connect with this researcher

Claiming links this public author record to a researcher profile and unlocks direct collaboration workflows.

Log in to claim

Direct collaboration

Open a focused conversation when the fit is right

Claim this author entity first to unlock direct invitations.

Research graph

See the researcher in context

Open full explorer

Inspect adjacent work, topics, institutions and collaborators without jumping out to a separate graph page.

Building this graph slice

BZPEER is loading the nearby papers, people, topics and institutions for this page.

Published work

15 published item(s)

preprint2026arXiv

Online Rack Placement in Large-Scale Data Centers: Online Sampling Optimization and Deployment

This paper optimizes the configuration of large-scale data centers toward cost-effective, reliable and sustainable cloud supply chains. The problem involves placing incoming racks of servers within a data center to maximize demand coverage given space, power and cooling restrictions. We formulate an online integer optimization model to support rack placement decisions. We propose a tractable online sampling optimization (OSO) approach to multi-stage stochastic optimization, which approximates unknown parameters with a sample path and re-optimizes decisions dynamically. We prove that OSO achieves a strong competitive ratio in canonical online resource allocation problems and sublinear regret in the online batched bin packing problem. Theoretical and computational results show it can outperform mean-based certainty-equivalent resolving heuristics. Our algorithm has been packaged into a software solution deployed across Microsoft's data centers, contributing an interactive decision-making process at the human-machine interface. Using deployment data, econometric tests suggest that adoption of the solution has a negative and statistically significant impact on power stranding, estimated at 1-3 percentage point. At the scale of cloud computing, these improvements in data center performance result in significant cost savings and environmental benefits.

0 citations0 reviewsNo code linkedOpen access
preprint2026arXiv

Online Scheduling for LLM Inference with KV Cache Constraints

Large Language Model (LLM) inference, where a trained model generates text one word at a time in response to user prompts, is a computationally intensive process requiring efficient scheduling to optimize latency and resource utilization. A key challenge in LLM inference is the management of the Key-Value (KV) cache, which reduces redundant computations but introduces memory constraints. In this work, we model LLM inference with KV cache constraints theoretically and propose a novel batching and scheduling algorithm that minimizes inference latency while effectively managing the KV cache's memory. More specifically, we make the following contributions. First, to evaluate the performance of online algorithms for scheduling in LLM inference, we introduce a hindsight optimal benchmark, formulated as an integer program that computes the minimum total inference latency under full future information. Second, we prove that no deterministic online algorithm can achieve a constant competitive ratio when the arrival process is arbitrary. Third, motivated by the computational intractability of solving the integer program at scale, we propose a polynomial-time online scheduling algorithm and show that under certain conditions it can achieve a constant competitive ratio. We also demonstrate our algorithm's strong empirical performance by comparing it to the hindsight optimal in a synthetic dataset. Finally, we conduct empirical evaluations on a real-world public LLM inference dataset, simulating the Llama2-70B model on A100 GPUs, and show that our algorithm significantly outperforms the benchmark algorithms. Overall, our results offer a path toward more sustainable and cost-effective LLM deployment.

0 citations0 reviewsNo code linkedOpen access
preprint2026arXiv

OptiMind: Teaching LLMs to Think Like Optimization Experts

Mathematical programming -- the task of expressing operations and decision-making problems in precise mathematical language -- is fundamental across domains, yet remains a skill-intensive process requiring operations research expertise. Recent advances in large language models for complex reasoning have spurred interest in automating this task, translating natural language into executable optimization models. Current approaches, however, achieve limited accuracy, hindered by scarce and noisy training data without leveraging domain knowledge. In this work, we systematically integrate optimization expertise to improve formulation accuracy for mixed-integer linear programming, a key family of mathematical programs. Our OptiMind framework leverages semi-automated, class-based error analysis to guide both training and inference, explicitly preventing common mistakes within each optimization class. Our resulting fine-tuned LLM significantly improves formulation accuracy by 20.7% across multiple optimization benchmarks, with consistent gains under test-time scaling methods such as self-consistency and multi-turn feedback, enabling further progress toward robust LLM-assisted optimization formulation.

0 citations0 reviewsNo code linkedOpen access
preprint2026arXiv

Sample Complexity of Stochastic Optimization with Integer Variables

We establish sample complexity results for stochastic optimization over the integers, especially with a view to understand the complexity with respect to the corresponding continuous optimization problem. We show that integer optimization can sometimes require strictly more samples and sometimes strictly smaller number of samples, depending on the structure of the objective and constraints. 1. For Lipschitz objectives over subsets of the $\ell_\infty$ ball, the statistical complexity of general stochastic mixed-integer, nonlinear, nonconvex optimization is exactly the same as stochastic linear optimization with just bound constraints. 2. For Lipschitz objectives over subsets of the $\ell_2$ ball, we show that integer optimization can require strictly *smaller* sample size compared to the continuous setting in a certain regime. To get to this result, we also establish tight sample complexity results for nonconvex continuous stochastic optimization which, to the best of our knowledge, do not appear in prior work. 3. For strongly convex, smooth objectives, integer optimization has high statistical complexity compared to the continuous setting. In particular, we show that integer optimization requires $Ω(1/ε^2)$ samples to report an $ε$-approximate solution, compared to the well-known $O(1/ε)$ sample complexity from the continuous optimization literature.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

IBIS-A: The IBIS data Archive. High resolution observations of the solar photosphere and chromosphere with contextual data

The IBIS data Archive (IBIS-A) stores data acquired with the Interferometric BIdimensional Spectropolarimeter (IBIS), which was operated at the Dunn Solar Telescope of the US National Solar Observatory from 2003 to 2019. The instrument provided series of high-resolution narrowband spectropolarimetric imaging observations of the photosphere and chromosphere in the range 5800$-$8600 Å~ and co-temporal broadband observations in the same spectral range and with the same field of view of the polarimetric data. We present the data currently stored in IBIS-A, as well as the interface utilized to explore such data and facilitate its scientific exploitation. To this purpose we also describe the use of IBIS-A data in recent and undergoing studies relevant to solar physics and space weather research. IBIS-A includes raw and calibrated observations, as well as science-ready data. The latter comprise maps of the circular, linear, and net circular polarization, and of the magnetic and velocity fields derived for a significant fraction of the series available in the archive. IBIS-A furthermore contains links to observations complementary to the IBIS data, such as co-temporal high-resolution observations of the solar atmosphere available from the instruments onboard the Hinode and IRIS satellites, and full-disc multiband images from INAF solar telescopes. IBIS-A currently consists of 30 TB of data taken with IBIS during 28 observing campaigns performed in 2008 and from 2012 to 2019 on 159 days. Metadata and movies of each calibrated and science-ready series are also available to help users evaluating observing conditions. IBIS-A represents a unique resource for investigating the plasma processes in the solar atmosphere and the solar origin of space weather events.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

Lipschitz Selectors may not Yield Competitive Algorithms for Convex Body Chasing

The current best algorithms for convex body chasing problem in online algorithms use the notion of the Steiner point of a convex set. In particular, the algorithm which always moves to the Steiner point of the request set is $O(d)$ competitive for nested convex body chasing, and this is optimal among memoryless algorithms [Bubeck et al. 2020]. A memoryless algorithm coincides with the notion of a selector in functional analysis. The Steiner point is noted for being Lipschitz with respect to the Hausdorff metric, and for achieving the minimal Lipschitz constant possible. It is natural to ask whether every selector with this Lipschitz property yields a competitive algorithm for nested convex body chasing. We answer this question in the negative by exhibiting a selector which yields a non-competitive algorithm for nested convex body chasing but is Lipschitz with respect to Hausdorff distance. Furthermore, we show that being Lipschitz with respect to an $L_p$-type analog to the Hausdorff distance is sufficient to guarantee competitiveness if and only if $p=1$.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

Lower Bounds on the Size of General Branch-and-Bound Trees

A \emph{general branch-and-bound tree} is a branch-and-bound tree which is allowed to use general disjunctions of the form $π^{\top} x \leq π_0 \,\vee\, π^{\top}x \geq π_0 + 1$, where $π$ is an integer vector and $π_0$ is an integer scalar, to create child nodes. We construct a packing instance, a set covering instance, and a Traveling Salesman Problem instance, such that any general branch-and-bound tree that solves these instances must be of exponential size. We also verify that an exponential lower bound on the size of general branch-and-bound trees persists when we add Gaussian noise to the coefficients of the cross polytope, thus showing that polynomial-size "smoothed analysis" upper bound is not possible. The results in this paper can be viewed as the branch-and-bound analog of the seminal paper by Chvátal et al. \cite{chvatal1989cutting}, who proved lower bounds for the Chvátal-Gomory rank.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

Online Demand Scheduling with Failovers

Motivated by cloud computing applications, we study the problem of how to optimally deploy new hardware subject to both power and robustness constraints. To model the situation observed in large-scale data centers, we introduce the Online Demand Scheduling with Failover problem. There are $m$ identical devices with capacity constraints. Demands come one-by-one and, to be robust against a device failure, need to be assigned to a pair of devices. When a device fails (in a failover scenario), each demand assigned to it is rerouted to its paired device (which may now run at increased capacity). The goal is to assign demands to the devices to maximize the total utilization subject to both the normal capacity constraints as well as these novel failover constraints. These latter constraints introduce new decision tradeoffs not present in classic assignment problems such as the Multiple Knapsack problem and AdWords. In the worst-case model, we design a deterministic $\approx \frac{1}{2}$-competitive algorithm, and show this is essentially tight. To circumvent this constant-factor loss, which in the context of big cloud providers represents substantial capital losses, we consider the stochastic arrival model, where all demands come i.i.d. from an unknown distribution. In this model we design an algorithm that achieves a sub-linear additive regret (i.e. as OPT or $m$ increases, the multiplicative competitive ratio goes to $1$). This requires a combination of different techniques, including a configuration LP with a non-trivial post-processing step and an online monotone matching procedure introduced by Rhee and Talagrand.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

Solving sparse principal component analysis with global support

Sparse principal component analysis with global support (SPCAgs), is the problem of finding the top-$r$ leading principal components such that all these principal components are linear combinations of a common subset of at most $k$ variables. SPCAgs is a popular dimension reduction tool in statistics that enhances interpretability compared to regular principal component analysis (PCA). Methods for solving SPCAgs in the literature are either greedy heuristics (in the special case of $r = 1$) with guarantees under restrictive statistical models or algorithms with stationary point convergence for some regularized reformulation of SPCAgs. Crucially, none of the existing computational methods can efficiently guarantee the quality of the solutions obtained by comparing them against dual bounds. In this work, we first propose a convex relaxation based on operator norms that provably approximates the feasible region of SPCAgs within a $c_1 + c_2 \sqrt{\log r} = O(\sqrt{\log r})$ factor for some constants $c_1, c_2$. To prove this result, we use a novel random sparsification procedure that uses the Pietsch-Grothendieck factorization theorem and may be of independent interest. We also propose a simpler relaxation that is second-order cone representable and gives a $(2\sqrt{r})$-approximation for the feasible region. Using these relaxations, we then propose a convex integer program that provides a dual bound for the optimal value of SPCAgs. Moreover, it also has worst-case guarantees: it is within a multiplicative/additive factor of the original optimal value, and the multiplicative factor is $O(\log r)$ or $O(r)$ depending on the relaxation used. Finally, we conduct computational experiments that show that our convex integer program provides, within a reasonable time, good upper bounds that are typically significantly better than the natural baselines.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

Time-Constrained Learning

Consider a scenario in which we have a huge labeled dataset ${\cal D}$ and a limited time to train some given learner using ${\cal D}$. Since we may not be able to use the whole dataset, how should we proceed? Questions of this nature motivate the definition of the Time-Constrained Learning Task (TCL): Given a dataset ${\cal D}$ sampled from an unknown distribution $μ$, a learner ${\cal L}$ and a time limit $T$, the goal is to obtain in at most $T$ units of time the classification model with highest possible accuracy w.r.t. to $μ$, among those that can be built by ${\cal L}$ using the dataset ${\cal D}$. We propose TCT, an algorithm for the TCL task designed based that on principles from Machine Teaching. We present an experimental study involving 5 different Learners and 20 datasets where we show that TCT consistently outperforms two other algorithms: the first is a Teacher for black-box learners proposed in [Dasgupta et al., ICML 19] and the second is a natural adaptation of random sampling for the TCL setting. We also compare TCT with Stochastic Gradient Descent training -- our method is again consistently better. While our work is primarily practical, we also show that a stripped-down version of TCT has provable guarantees. Under reasonable assumptions, the time our algorithm takes to achieve a certain accuracy is never much bigger than the time it takes the batch teacher (which sends a single batch of examples) to achieve similar accuracy, and in some case it is almost exponentially better.

0 citations0 reviewsNo code linkedOpen access
preprint2021arXiv

Spreading the word -- current status of VO tutorials and schools

With some telescopes standing still, now more than ever simple access to archival data is vital for astronomers and they need to know how to go about it. Within European Virtual Observatory (VO) projects, such as AIDA (2008-2010), ICE (2010-2012), CoSADIE (2013-2015), ASTERICS (2015-2018) and ESCAPE (since 2019), we have been offering Virtual Observatory schools for many years. The aim of these schools are twofold: teaching (early career) researchers about the functionalities and possibilities within the Virtual Observatory and collecting feedback from the astronomical community. In addition to the VO schools on the European level, different national teams have also put effort into VO dissemination. The team at the Centre de Données astronomiques de Strasbourg (CDS) started to explore more and new ways to interact with the community: a series of blog posts on AstroBetter.com or a lunch time session at the virtual EAS meeting 2020. The Spanish VO has conducted virtual VO schools. GAVO has supported online archive workshops and maintains their Virtual Observatory Text Treasures. In this paper, we present the different formats in more detail, and report on the resulting interaction with the community as well as the estimated reach.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

Exo-MerCat: a merged exoplanet catalog with Virtual Observatory connection

The heterogeneity of papers dealing with the discovery and characterization of exoplanets makes every attempt to maintain a uniform exoplanet catalog almost impossible. Four sources currently available online (NASA Exoplanet Archive, Exoplanet Orbit Database, Exoplanet Encyclopaedia, and Open Exoplanet Catalogue) are commonly used by the community, but they can hardly be compared, due to discrepancies in notations and selection criteria. Exo-MerCat is a Python code that collects and selects the most precise measurement for all interesting planetary and orbital parameters contained in the four databases, accounting for the presence of multiple aliases for the same target. It can download information about the host star as well by the use of Virtual Observatory ConeSearch connections to the major archives such as SIMBAD and those available in VizieR. A Graphical User Interface is provided to filter data based on the user's constraints and generate automatic plots that are commonly used in the exoplanetary community. With Exo-MerCat, we retrieved a unique catalog that merges information from the four main databases, standardizing the output and handling notation differences issues. Exo-MerCat can correct as many issues that prevent a direct correspondence between multiple items in the four databases as possible, with the available data. The catalog is available as a VO resource for everyone to use and it is periodically updated, according to the update rates of the source catalogs.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

Knapsack Secretary with Bursty Adversary

The random-order or secretary model is one of the most popular beyond-worst case model for online algorithms. While it avoids the pessimism of the traditional adversarial model, in practice we cannot expect the input to be presented in perfectly random order. This has motivated research on ``best of both worlds'' (algorithms with good performance on both purely stochastic and purely adversarial inputs), or even better, on inputs that are a mix of both stochastic and adversarial parts. Unfortunately the latter seems much harder to achieve and very few results of this type are known. Towards advancing our understanding of designing such robust algorithms, we propose a random-order model with bursts of adversarial time steps. The assumption of burstiness of unexpected patterns is reasonable in many contexts, since changes (e.g. spike in a demand for a good) are often triggered by a common external event. We then consider the Knapsack Secretary problem in this model: there is a knapsack of size $k$ (e.g., available quantity of a good), and in each of the $n$ time steps an item comes with its value and size in $[0,1]$ and the algorithm needs to make an irrevocable decision whether to accept or reject the item. We design an algorithm that gives an approximation of $1 - \tilde{O}(Γ/k)$ when the adversarial time steps can be covered by $Γ\ge \sqrt{k}$ intervals of size $\tilde{O}(\frac{n}{k})$. In particular, setting $Γ= \sqrt{k}$ gives a $(1 - O(\frac{\ln^2 k}{\sqrt{k}}))$-approximation that is resistant to up to a $\frac{\ln^2 k}{\sqrt{k}}$-fraction of the items being adversarial, which is almost optimal even in the absence of adversarial items. Also, setting $Γ= \tildeΩ(k)$ gives a constant approximation that is resistant to up to a constant fraction of items being adversarial.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

Sparse PSD approximation of the PSD cone

While semidefinite programming (SDP) problems are polynomially solvable in theory, it is often difficult to solve large SDP instances in practice. One technique to address this issue is to relax the global positive-semidefiniteness (PSD) constraint and only enforce PSD-ness on smaller $k\times k$ principal submatrices --- we call this the sparse SDP relaxation. Surprisingly, it has been observed empirically that in some cases this approach appears to produce bounds that are close to the optimal objective function value of the original SDP. In this paper, we formally attempt to compare the strength of the sparse SDP relaxation vis-à-vis the original SDP from a theoretical perspective. In order to simplify the question, we arrive at a data independent version of it, where we compare the sizes of SDP cone and the $k$-PSD closure, which is the cone of matrices where PSD-ness is enforced on all $k\times k$ principal submatrices. In particular, we investigate the question of how far a matrix of unit Frobenius norm in the $k$-PSD closure can be from the SDP cone. We provide two incomparable upper bounds on this farthest distance as a function of $k$ and $n$. We also provide matching lower bounds, which show that the upper bounds are tight within a constant in different regimes of $k$ and $n$. Other than linear algebra techniques, we extensively use probabilistic methods to arrive at these bounds. One of the lower bounds is obtained by observing a connection between matrices in the $k$-PSD closure and matrices satisfying the restricted isometry property (RIP).

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

The GAPS Programme at TNG XXVIII -- A pair of hot-Neptunes orbiting the young star TOI-942

Both young stars and multi-planet systems are primary objects that allow us to study, understand and constrain planetary formation and evolution theories. We validate the physical nature of two Neptune-type planets transiting TOI-942 (TYC 5909-319-1), a previously unacknowledged young star (50+30-20 Myr) observed by the TESS space mission in Sector 5. Thanks to a comprehensive stellar characterization, TESS light curve modelling and precise radial-velocity measurements, we validated the planetary nature of the TESS candidate and detect an additional transiting planet in the system on a larger orbit. From photometric and spectroscopic observations we performed an exhaustive stellar characterization and derived the main stellar parameters. TOI-942 is a relatively active K2.5V star (logR'hk = -4.17+-0.01) with rotation period Prot = 3.39+-0.01 days, a projected rotation velocity vsini=13.8+-0.5 km/s and a radius of ~0.9 Rsun. We found that the inner planet, TOI-942b, has an orbital period Pb=4.3263+-0.0011 days, a radius Rb=4.242-0.313+0.376 Rearth and a mass upper limit of 16 Mearth at 1-sigma confidence level. The outer planet, TOI-942c, has an orbital period Pc=10.1605-0.0053+0.0056 days, a radius Rc=4.793-0.351+0.410 Rearth and a mass upper limit of 37 Mearth at 1-sigma confidence level.

0 citations0 reviewsNo code linkedOpen access