arXiv subjects

Bala Krishnamoorthy

Publications and source records attributed to Bala Krishnamoorthy.

At least 19 recordsLinked to original sources

arXiv2026

Existence of Minimal Homotopies for Immersed Planar Curves

We study the existence of area-minimizing homotopies between homotopic curves in the plane. While the classical Plateau problem establishes the existence of least-area surfaces spanning a single Jordan curve, the corresponding existence theory for homotopies between curves is more subtle and is not directly covered by the same framework. Existing results in the plane are mainly based on combinatorial and algebraic methods, such as decomposing curves into self-overlapping subcurves. These methods are highly effective in the planar setting, but they are often tied to special classes of curves and rely strongly on the local structure of the self-intersections, frequently assuming transverse crossings. In contrast, our approach is geometric and variational, and does not depend on the local structure of the self-intersections. In this paper, we develop a variational existence theory for minimum-area homotopies of immersed planar curves. Our approach adapts classical minimal surface methods by lifting an immersed planar curve with self-intersections into higher co-dimension, where it becomes embedded. For such a lifted curve, we apply Douglas's solution of the Plateau problem to obtain an area-minimizing disk. For closed curves of class $C^1$, we prove uniform convergence of the Douglas minimizers and show that the limiting map minimizes area among all $C^1$ spanning maps of the original planar curve. We then extend the construction to closed Lipschitz curves using approximation and Sobolev compactness arguments. Since the limiting minimizing disk lies in the original plane, it directly produces a null homotopy whose swept area is minimal among all admissible homotopies of the original curve. In this way, the construction connects Plateau theory with minimal homotopy area minimization.

math.GT↗

arXiv2026

Minimal Homotopies in Three Dimensions: A Cable System Approach

We study null homotopies of immersed spheres in $\mathbb{R}^3$ and the volume they sweep during contraction. For a smooth immersion with finitely many transverse self-intersections, we introduce a cable system that connects each bounded region of the complement to the exterior. From this construction we define the cable index and prove that it agrees with the Brouwer degree on each complementary region. Using this identification, we derive a degree-weighted lower bound for the swept volume of any Lipschitz null homotopy. We show that the bound is attained whenever the homotopy is sense-preserving, meaning the surface moves in a consistent direction, and the index evolves monotonically along the homotopy. In addition, in the case where the immersion arises as the boundary of an immersed ball, we construct an explicit homotopy that realizes this lower bound via a deformation of the ball. Finally, we present a linear-time algorithm that computes all cable indices from a finite cable system, providing a concrete and computable method for evaluating the lower bound.

math.GT↗

arXiv2026

A Stable Measure of Chaos in Dynamical Systems using Persistent Homology

Many real-world dynamics exhibit chaos, a phenomenon in which neighboring trajectories in the state space of a dynamical system diverge exponentially over time. A common measure used for quantifying the degree of this divergence is the maximal Lyapunov exponent, which relies on pairwise Euclidean distances between the trajectories at each time. The main limitation of the maximal Lyapunov exponent in practice is its sensitivity to small perturbations in system trajectories. Persistent homology, the study of holes that appear in different dimensions as the points of a data set are thickened over time, has guaranteed theoretical stability under such added noise. As such, we propose a novel, 0-dimensional persistent homology based measure of chaos termed the 0-persistence exponent and prove its theoretical stability. We show that if a system is chaotic, then the 0-persistence exponent is non-negative by proving that positive Lyapunov measures imply non-negative 0-persistence measures, and further discuss when strict positivity of 0-persistence measures occur. Additionally, we prove the existence of an upper bound on our measure, and show its greater experimental stability on the Lorenz and Rossler systems describing fluid convection and taffy pulling. We present an algorithm for computing the 0-persistence exponent given a single univariate time series with N points from a dynamical system that runs in O(N^2 log N) time. We finally show the greater experimental stability of the 0-persistence exponent on time series data depicting a Belousov-Zhabotinsky chemical reaction, which transitions from periodicity to chaos and back as the system evolves in time. We present experimental results which verify that positive Lyapunov exponents imply positive 0-persistence exponents under sufficient conditions through high correlation between both measures on the Lorenz and Rossler systems.

math.AT↗

arXiv2025

Weighted MCC: A Robust Measure of Multiclass Classifier Performance for Observations with Individual Weights

Several performance measures are used to evaluate binary and multiclass classification tasks. But individual observations may often have distinct weights, and none of these measures are sensitive to such varying weights. We propose a new weighted Pearson-Matthews Correlation Coefficient (MCC) for binary classification as well as weighted versions of related multiclass measures. The weighted MCC varies between $-1$ and $1$. But crucially, the weighted MCC values are higher for classifiers that perform better on highly weighted observations, and hence is able to distinguish them from classifiers that have a similar overall performance and ones that perform better on the lowly weighted observations. Furthermore, we prove that the weighted measures are robust with respect to the choice of weights in a precise manner: if the weights are changed by at most $\epsilon$, the value of the weighted measure changes at most by a factor of $\epsilon$ in the binary case and by a factor of $\epsilon^2$ in the multiclass case. Our computations demonstrate that the weighted measures clearly identify classifiers that perform better on higher weighted observations, while the unweighted measures remain completely indifferent to the choices of weights.

stat.ML↗

arXiv2025

Dynamic Learning Rate Scheduling based on Loss Changes Leads to Faster Convergence

Despite significant advances in optimizers for training, most research works use common scheduler choices like Cosine or exponential decay. In this paper, we study \emph{GreedyLR}, a novel scheduler that adaptively adjusts the learning rate during training based on the current loss. To validate the effectiveness of our proposed scheduler, we conduct experiments on several NLP, CV, and LLM tasks with up to $7B$ parameters, including both fine-tuning and pre-training experiments. The results show that our approach outperforms several state-of-the-art schedulers in terms of accuracy, speed, and convergence. We also provide a theoretical analysis of the GreedyLR algorithm, including a proof of convergence and derivation of the optimal scaling factor $F$ that maximizes the convergence rate, along with experiments to show robustness of the algorithm to realistic noisy landscapes. Our scheduler is easy to implement, computationally efficient, and could be considered a good default scheduler for training.

cs.AI↗

arXiv2025

Scalable Two-Stage Stochastic Optimal Power Flow via Separable Approximation

This paper proposes a Separable Projective Approximation Routine-Optimal Power Flow (SPAR-OPF) framework for solving two-stage stochastic optimization problems in power systems. The framework utilizes a separable piecewise linear approximation of the value function and learns the function based on sample sub-gradient information. We present two formulations to model the learned value function, and compare their effectiveness. Additionally, an efficient statistical method is introduced to assess the quality of the obtained solutions. The effectiveness of the proposed framework is validated using distributed generation siting and sizing problem in three-phase unbalanced power distribution systems as an example. Results show that the framework approximates the value function with over 98% accuracy and provides high-quality solutions with an optimality gap of less than 1%. The framework scales efficiently with system size, generating high-quality solutions in a short time when applied to a 9500-node distribution system with 1200 scenarios, while the extensive formulations and progressive hedging failed to solve the problem.

eess.SY↗

arXiv2025

A Stable Measure of Similarity for Time Series using Persistent Homology

Persistent homology, the study of holes that appear in data as one thickens balls centered around its points over time, has theoretically guaranteed stability. That is, small data perturbations guarantee small changes in the lifetimes of these holes. This stability has been used to construct a measure of periodicity for a single univariate time series, denoted score(f1). One popular measure of similarity between two time series is percent determinism (%DET), which measures the correlation between two time-series embeddings. We introduce a novel persistent-homology based measure of time-series similarity which we denote the bi-conditional periodicity score, score(f1,f2). We prove the stability of our measure under small time series and frequency perturbations, as well as the existence of a minimum embedding dimension for the convergence of our score. Our latter result implies that larger embedding dimensions may be necessary to reach desired levels of convergence. Since pairwise distances between points in these larger dimensions may start to concentrate, we also prove the stability of our measure under dimension reduction which guarantees that as long as the first K principal components capture a majority of the variance under orthogonal projection, the score will undergo small changes. We next introduce an algorithm for computing the bi-conditional periodicity score and deduce its computational complexity as O(N log N + PK^2 + P^6) for N the number of time series points, P the number of embedding points, and K the number of principal components. We experimentally verify the greater stability of our measure in comparison with %DET on both synthetic time series as well as real climate data. As well, score(f1,f2) requires only one parameter for its computation while %DET requires four.

math.AT↗

arXiv2024

Learning to Simulate Aerosol Dynamics with Graph Neural Networks

Aerosol effects on climate, weather, and air quality depend on characteristics of individual particles, which are tremendously diverse and change in time. Particle-resolved models are the only models able to capture this diversity in particle physiochemical properties, and these models are computationally expensive. As a strategy for accelerating particle-resolved microphysics models, we introduce Graph-based Learning of Aerosol Dynamics (GLAD) and use this model to train a surrogate of the particle-resolved model PartMC-MOSAIC. GLAD implements a Graph Network-based Simulator (GNS), a machine learning framework that has been used to simulate particle-based fluid dynamics models. In GLAD, each particle is represented as a node in a graph, and the evolution of the particle population over time is simulated through learned message passing. We demonstrate our GNS approach on a simple aerosol system that includes condensation of sulfuric acid onto particles composed of sulfate, black carbon, organic carbon, and water. A graph with particles as nodes is constructed, and a graph neural network (GNN) is then trained using the model output from PartMC-MOSAIC. The trained GNN can then be used for simulating and predicting aerosol dynamics over time. Results demonstrate the framework's ability to accurately learn chemical dynamics and generalize across different scenarios, achieving efficient training and prediction times. We evaluate the performance across three scenarios, highlighting the framework's robustness and adaptability in modeling aerosol microphysics and chemistry.

physics.ao-ph↗

arXiv2024

Box Filtration

We define a new framework that unifies the filtration and mapper approaches from TDA, and present efficient algorithms to compute it. Termed the box filtration of a PCD, we grow boxes (hyperrectangles) that are not necessarily centered at each point (in place of balls centered at points). We grow the boxes non-uniformly and asymmetrically in different dimensions based on the distribution of points. We present two approaches to handle the boxes: a point cover where each point is assigned its own box at start, and a pixel cover that works with a pixelization of the space of the PCD. Any box cover in either setting automatically gives a mapper of the PCD. We show that the persistence diagrams generated by the box filtration using both point and pixel covers satisfy the classical stability based on the Gromov-Hausdorff distance. Using boxes also implies that the box filtration is identical for pairwise or higher order intersections whereas the VR and Cech filtration are not the same. Growth in each dimension is computed by solving a linear program (LP) that optimizes a cost functional balancing the cost of expansion and benefit of including more points in the box. The box filtration algorithm runs in $O(m|U(0)|\log(mn\pi)L(q))$ time, where $m$ is number of steps of increments considered for box growth, $|U(0)|$ is the number of boxes in the initial cover ($\leq$ number of points), $\pi$ is the step length for increasing each box dimension, each LP is solved in $O(L(q))$ time, $n$ is the PCD dimension, and $q = n \times |X|$. We demonstrate through multiple examples that the box filtration can produce more accurate results to summarize the topology of the PCD than VR and distance-to-measure (DTM) filtrations. Software for our implementation is available at https://github.com/pragup/Box-Filteration.

cs.CG↗

arXiv2024

Predictive Analytics of Varieties of Potatoes

We explore the application of machine learning algorithms specifically to enhance the selection process of Russet potato clones in breeding trials by predicting their suitability for advancement. This study addresses the challenge of efficiently identifying high-yield, disease-resistant, and climate-resilient potato varieties that meet processing industry standards. Leveraging manually collected data from trials in the state of Oregon, we investigate the potential of a wide variety of state-of-the-art binary classification models. The dataset includes 1086 clones, with data on 38 attributes recorded for each clone, focusing on yield, size, appearance, and frying characteristics, with several control varieties planted consistently across four Oregon regions from 2013-2021. We conduct a comprehensive analysis of the dataset that includes preprocessing, feature engineering, and imputation to address missing values. We focus on several key metrics such as accuracy, F1-score, and Matthews correlation coefficient (MCC) for model evaluation. The top-performing models, namely a neural network classifier (Neural Net), histogram-based gradient boosting classifier (HGBC), and a support vector machine classifier (SVM), demonstrate consistent and significant results. To further validate our findings, we conduct a simulation study. By simulating different data-generating scenarios, we assess model robustness and performance through true positive, true negative, false positive, and false negative distributions, area under the receiver operating characteristic curve (AUC-ROC) and MCC. The simulation results highlight that non-linear models like SVM and HGBC consistently show higher AUC-ROC and MCC than logistic regression (LR), thus outperforming the traditional linear model across various distributions, and emphasizing the importance of model selection and tuning in agricultural trials.

cs.LG↗

arXiv2024

Structural Validation Of Synthetic Power Distribution Networks Using The Multiscale Flat Norm

We study the problem of comparing a pair of geometric networks that may not be similarly defined, i.e., when they do not have one-to-one correspondences between their nodes and edges. Our motivating application is to compare power distribution networks of a region. Due to the lack of openly available power network datasets, researchers synthesize realistic networks resembling their actual counterparts. But the synthetic digital twins may vary significantly from one another and from actual networks due to varying underlying assumptions and approaches. Hence the user wants to evaluate the quality of networks in terms of their structural similarity to actual power networks. But the lack of correspondence between the networks renders most standard approaches, e.g., subgraph isomorphism and edit distance, unsuitable. We propose an approach based on the multiscale flat norm, a notion of distance between objects defined in the field of geometric measure theory, to compute the distance between a pair of planar geometric networks. Using a triangulation of the domain containing the input networks, the flat norm distance between two networks at a given scale can be computed by solving a linear program. In addition, this computation automatically identifies the 2D regions (patches) that capture where the two networks are different. We demonstrate through 2D examples that the flat norm distance can capture the variations of inputs more accurately than the commonly used Hausdorff distance. As a notion of stability, we also derive upper bounds on the flat norm distance between a simple 1D curve and its perturbed version as a function of the radius of perturbation for a restricted class of perturbations. We demonstrate our approach on a set of actual power networks from a county in the USA.

cs.CG↗

arXiv2024

Approximate Bipartite $b$-Matching using Multiplicative Auction

Given a bipartite graph $G(V= (A \cup B),E)$ with $n$ vertices and $m$ edges and a function $b \colon V \to \mathbb{Z}_+$, a $b$-matching is a subset of edges such that every vertex $v \in V$ is incident to at most $b(v)$ edges in the subset. When we are also given edge weights, the Max Weight $b$-Matching problem is to find a $b$-matching of maximum weight, which is a fundamental combinatorial optimization problem with many applications. Extending on the recent work of Zheng and Henzinger (IPCO, 2023) on standard bipartite matching problems, we develop a simple auction algorithm to approximately solve Max Weight $b$-Matching. Specifically, we present a multiplicative auction algorithm that gives a $(1 - \varepsilon)$-approximation in $O(m \varepsilon^{-1} \log \varepsilon^{-1} \log \beta)$ worst case time, where $\beta$ the maximum $b$-value. Although this is a $\log \beta$ factor greater than the current best approximation algorithm by Huang and Pettie (Algorithmica, 2022), it is considerably simpler to present, analyze, and implement.

cs.DS↗

arXiv2023

Semi-Streaming Algorithms for Weighted $k$-Disjoint Matchings

We design and implement two single-pass semi-streaming algorithms for the maximum weight $k$-disjoint matching ($k$-DM) problem. Given an integer $k$, the $k$-DM problem is to find $k$ pairwise edge-disjoint matchings such that the sum of the weights of the matchings is maximized. For $k \geq 2$, this problem is NP-hard. Our first algorithm is based on the primal-dual framework of a linear programming relaxation of the problem and is $\frac{1}{3+\varepsilon}$-approximate. We also develop an approximation preserving reduction from $k$-DM to the maximum weight $b$-matching problem. Leveraging this reduction and an existing semi-streaming $b$-matching algorithm, we design a $(\frac{1}{2+\varepsilon})(1 - \frac{1}{k+1})$-approximate semi-streaming algorithm for $k$-DM. For any constant $\varepsilon > 0$, both of these algorithms require $O(nk \log_{1+\varepsilon}^2 n)$ bits of space. To the best of our knowledge, this is the first study of semi-streaming algorithms for the $k$-DM problem. We compare our two algorithms to state-of-the-art offline algorithms on 95 real-world and synthetic test problems, including thirteen graphs generated from data center network traces. On these instances, our streaming algorithms used significantly less memory (ranging from 6$\times$ to 512$\times$ less) and were faster in runtime than the offline algorithms. Our solutions were often within 5% of the best weights from the offline algorithms. We highlight that the existing offline algorithms run out of 1 TB memory for most of the large instances ($>1$ billion edges), whereas our streaming algorithms can solve these problems using only 100 GB memory for $k=8$.

cs.DS↗

arXiv2023

Geometry of a Set and its Random covers

Let $E$ be a bounded open subset of $\mathbb{R}^n$. We study the following questions: For i.i.d. samples $X_1, \dots, X_N$ drawn uniformly from $E$, what is the probability that $\cup_i \mathbf{B}(X_i, δ)$, the union of $δ$-balls centered at $X_i$, covers $E$? And how does the probability depend on sample size $N$ and the radius of balls $δ$? We present geometric conditions of $E$ under which we derive lower bounds to this probability. These lower bounds tend to $1$ as a function of $\exp{(-δ^n N)}$. The basic tool that we use to derive the lower bounds is a good partition of $E$, i.e., one whose partition elements have diameters that are uniformly bounded from above and have volumes that are uniformly bounded from below. We show that if $E^c$, the complement of $E$, has positive reach then we can construct a good partition of $E$. This partition is motivated by the Whitney decomposition of $E$. On the other hand, we identify a class of bounded open subsets of $\mathbb{R}^n$ that do not satisfy this positive reach condition but do have good partitions. In 2D when $E^c\subset \mathbb{R}^2$ does not have positive reach, we show that the mutliscale flat norm can be used to approximate $E$ with a set that has a good partition under certain conditions. In this case, we provide a lower bound on the probability that the union of the balls almost covers $E$.

math.PR↗

arXiv2023

Convergence Guarantees of a Distributed Network Equivalence Algorithm for Distribution-OPF

The massive integration of distributed energy resources changes the operational demands of the electric power distribution system, motivating optimization-based approaches. The added computational complexities of the resulting optimal power flow (OPF) problem have generally been managed by approximated or relaxed models; however, they may lead to infeasible or inaccurate solutions. Decomposition-based methods have also been used to solve the OPF problems. But the existing methods require several message passing rounds for relatively small systems, causing significant delays in decision making; related feedback-based methods also suffer from slow tracking of the optimal solutions. In this paper, we propose a provably convergent distributed algorithm to solve the nonlinear OPF problem for power distribution systems. Our method is based on a previously developed decomposition-based optimization method that employs the network equivalence method. We present a thorough mathematical analysis that includes sufficient conditions that guarantee convergence of the method. We also present simulation results using the IEEE-123 bus test system to demonstrate the algorithm's effectiveness and provide additional insights into theoretical results.

math.OC↗

arXiv2023

A Normalized Bottleneck Distance on Persistence Diagrams and Homology Preservation under Dimension Reduction

Persistence diagrams (PDs) are used as signatures of point cloud data. Two clouds of points can be compared using the bottleneck distance d_B between their PDs. A potential drawback of this pipeline is that point clouds sampled from topologically similar manifolds can have arbitrarily large d_B when there is a large scaling between them. This situation is typical in dimension reduction frameworks. We define, and study properties of, a new scale-invariant distance between PDs termed normalized bottleneck distance, d_N. In defining d_N, we develop a broader framework called metric decomposition for comparing finite metric spaces of equal cardinality with a bijection. We utilize metric decomposition to prove a stability result for d_N by deriving an explicit bound on the distortion of the bijective map. We then study two popular dimension reduction techniques, Johnson-Lindenstrauss (JL) projections and metric multidimensional scaling (mMDS), and a third class of general biLipschitz mappings. We provide new bounds on how well these dimension reduction techniques preserve homology with respect to d_N. For a JL map f that transforms input X to f(X), we show that d_N(dgm(X),dgm(f(X))) < e, where dgm(X) is the Vietoris-Rips PD of X, and pairwise distances are preserved by f up to the tolerance 0 < \epsilon < 1. For mMDS, we present new bounds for d_B and d_N between PDs of X and its projection in terms of the eigenvalues of the covariance matrix. And for k-biLipschitz maps, we show that d_N is bounded by the product of (k^2-1)/k and the ratio of diameters of X and f(X). Finally, we use computational experiments to demonstrate the increased effectiveness of using the normalized bottleneck distance for clustering sets of point clouds sampled from different shapes.

cs.CG↗

arXiv2023

Robust Feasibility of Systems of Quadratic Equations Using Topological Degree Theory

We consider the problem of measuring the margin of robust feasibility of solutions to a system of nonlinear equations. We study the special case of a system of quadratic equations, which shows up in many practical applications such as the power grid and other infrastructure networks. This problem is a generalization of quadratically constrained quadratic programming (QCQP), which is NP-Hard in the general setting. We develop approaches based on topological degree theory to estimate bounds on the robustness margin of such systems. Our methods use tools from convex analysis and optimization theory to cast the problems of checking the conditions for robust feasibility as a nonlinear optimization problem. We then develop inner bound and outer bound procedures for this optimization problem, which could be solved efficiently to derive lower and upper bounds, respectively, for the margin of robust feasibility. We evaluate our approach numerically on standard instances taken from the MATPOWER and NESTA databases of AC power flow equations that describe the steady state of the power grid. The results demonstrate that our approach can produce tight lower and upper bounds on the margin of robust feasibility for such instances.

math.OC↗

arXiv2023

Persistent Homology to Study Cold Hardiness of Grape Cultivars

Persistent homology is a branch of computational algebraic topology that studies shapes and extracts features over multiple scales. In this paper, we present an unsupervised approach that uses persistent homology to study divergent behavior in agricultural point cloud data. More specifically, we build persistence diagrams from multidimensional point clouds, and use those diagrams as the basis to compare and contrast different subgroups of the population. We apply the framework to study the cold hardiness behavior of 5 leading grape cultivars, with real data from over 20 growing seasons. Our results demonstrate that persistent homology is able to effectively elucidate divergent behavior among the different cultivars; identify cultivars that exhibit variable behavior across seasons; and identify seasonal correlations.

math.AT↗