Simple Algorithm Research Articles

Respecting the Lane: A Lane Departure Identification for ADAS in Various Environmental and Lighting Conditions

This paper presents a novel methodology for estimating vehicle lane deviation in driving and issuing timely alerts, in varying environmental lighting conditions such as night, day and rainy, while making Advanced Driver Assistance Systems (ADAS) improved and more accessible to all users. A Fully Convolutional Neural Network (FCNN), trained on dashboard camera video, is utilized for semantic segmentation to identify the drivable area under varying lighting and weather conditions. This is followed by lane deviation computation using a novel geometrical algorithm, ultimately producing lane deviation warning. Additionally, a U-Net model is integrated for generating lane lines in scenarios where visibility is impaired or lane markings are unclear. The proposed system has shown remarkably low inference times compared to existing models with similar accuracy in lane recognition. The departure prediction has shown exceptional specificity and negative predicted value. The proposed system, which uses a simplified algorithms and low-resolution inputs, is designed for low computational and resource deployments and can adapt to rainy, daytime, and nighttime environments. The proposed system has promising potential to strengthen road safety by providing drivers with a readily deployable ADAS, tailored for vehicles that do not have manufacturer-installed systems. Integrating real-time traffic and road condition information could enable proactive adjustments to driving behaviour, potentially avoiding accidents and paving the way for safer and more intelligent transportation systems.

A Comprehensive Simplified Algorithm for Heat Transfer Modeling of Medium-Deep Borehole Heat Exchangers Considering Soil Stratification and Geothermal Gradient

Medium-deep borehole heat exchanger (BHE) systems represent an emerging form of ground source heat pump technology. Their heat transfer process is significantly influenced by geothermal gradient and soil stratification, typically simulated using segmented finite line source (SFLS) models. However, this approach involves computationally intensive procedures that hinder practical engineering implementation. Building upon an SFLS model adapted for complex geological conditions, this study proposes a comprehensive simplified algorithm: (1) For soil stratification: A geothermally-weighted thermal conductivity method converts layered heterogeneous media into an equivalent homogeneous medium; (2) For geothermal gradient: A temperature correction method establishes fluid temperatures under geothermal gradient by superimposing correction terms onto uniform-temperature model results (g-function model). Validated through two engineering case studies, this integrated algorithm provides a straightforward technical tool for heat transfer calculations in BHE systems.

AI-enhanced behavioral approach to measuring hearing in infants and toddlers: Proof-of-Concept Study.

Show that a basic unsupervised machine-learning algorithm can give information on whether a child reacted to a sound using facial, non-identifyable features recorded with a camera. Infants and toddlers were presented warble tones or single-syllable utterances 45 degrees to the left or to the right. A camera recorded their reactions, from which features like head turns or eye gaze were extracted with OpenFace. Three clusters were formed using Expectation Maximization on 80% of the toddler data. The remaining 20% and the infant data were used to verify if the clusters represent groups for sound presentations to the left, to the right and both directions. 28 infants (2-5 months) and 30 toddlers (2-4 years) were presented ten sounds each. The largest cluster comprised 90% of the trials with sound presentations in both directions, indicating "no decision". The remaining two clusters could be interpreted to represent reactions to the left and the right, respectively, and average sensitivities of 96% for the toddlers and 68% for the infants. A simple machine-learning algorithm demonstrated that it can form correct decisions on the direction of sound presentation by using anonymous facial data.

Predictions for the Detectability of Milky Way Satellite Galaxies and Outer-Halo Star Clusters with the Vera C. Rubin Observatory

We predict the sensitivity of the Vera C. Rubin Observatory Legacy Survey of Space and Time (LSST) to faint, resolved Milky Way satellite galaxies and outer-halo star clusters. We characterize the expected sensitivity using simulated LSST data from the LSST Dark Energy Science Collaboration (DESC) Data Challenge 2 (DC2) accessed and analyzed with the Rubin Science Platform as part of the Rubin Early Science Program. We simulate resolved stellar populations of Milky Way satellite galaxies and outer-halo star clusters over a wide range of sizes, luminosities, and heliocentric distances, which are broadly consistent with expectations for the Milky Way satellite system. We inject simulated stars into the DC2 catalog with realistic photometric uncertainties and star/galaxy separation derived from the DC2 data itself. We assess the probability that each simulated system would be detected by LSST using a conventional isochrone matched-filter technique. We find that assuming perfect star/galaxy separation enables the detection of resolved stellar systems with MV = 0 mag and r1/2 = 10 pc with >50% efficiency out to a heliocentric distance of ~250 kpc. Similar detection efficiency is possible with a simple star/galaxy separation criterion based on measured quantities, although the false positive rate is higher due to leakage of background galaxies into the stellar sample. When assuming perfect star/galaxy classification and a model for the galaxy-halo connection fit to current data, we predict that 89 +/- 20 Milky Way satellite galaxies will be detectable with a simple matched-filter algorithm applied to the LSST wide-fast-deep data set. Different assumptions about the performance of star/galaxy classification efficiency can decrease this estimate by ~7-25%, which emphasizes the importance of high-quality star/galaxy separation for studies of the Milky Way satellite population with LSST.

Research on self-supervised super resolution restoration algorithm based on reflective micro-scanning optical system

The features of infrared image such as less details and low SNR become bottleneck of infrared image application. This paper mainly focuses on research of super-resolution restoration algorithm of infrared image based on reflective infrared micro-scanning optical system. Aiming at solving super-resolution restoration problem of infrared image, self-supervised super resolution restoration algorithm is proposed and optimized. Meanwhile, reflective infrared micro-scanning optical system is introduced to break theoretical limit of simple image processing algorithm. And performance of infrared image super-resolution restoration is improved.

Temporal ghost imaging for pump-probe X-ray solution scattering.

Time-resolved small- and wide-angle X-ray scattering is a valuable tool for investigating biomolecular dynamics on a wide variety of timescales, without cryo-freezing or crystallization. However, some systems, such as the initial excitation of photo-active proteins, evolve dynamically on timescales that may be faster than the duration of the pump and probe beams. Data from a single pump-probe pulse pair therefore contain information from a mixture of time points. In this work, a simple algorithm is developed to recover the dynamics of solution scattering profiles. It leverages information about the pump and probe pulse beams' temporal profiles by using the same mathematical framework as ghost imaging [Pittman et al. (1995). Phys. Rev. A 52, R3429-R3432; Bennink et al. (2002). Phys. Rev. Lett. 89, 113601; Gatti et al. (2004). Phys. Rev. Lett. 93, 093602]. Results from several simulated data sets are presented.

Simpler and Faster Pairings from the Montgomery Ladder

We show that Montgomery ladders compute pairings as a by-product, and explain how a small adjustment to the ladder results in simple and efficient algorithms for the Weil and Tate pairing on elliptic curves using cubical arithmetic. We demonstrate the efficiency of the resulting cubical pairings in several applications from isogeny-based cryptography. Cubical pairings are simpler and more performant than pairings computed using Miller's algorithm: we get a speed-up of over 40 per cent for use-cases in SQIsign, and a speed-up of about 7 per cent for use-cases in CSIDH. While these results arise from a deep connection to biextensions and cubical arithmetic, in this article we keep things as concrete (and digestible) as possible. We provide a concise and complete introduction to cubical arithmetic as an appendix.

Optimization of Arrangements of Heat-Storage Bricks in a Regenerative Combustion System by Tree Search

When there are several different types of heat-storage ceramic bricks (checkers) that can be arranged in a regenerative combustion system, one must find an optimal arrangement (with the highest long-term Waste Heat Recovery Ratio) of these checkers, possibly of different types, in this regenerative combustion system. However, the number of possible arrangements of checkers in a heat regenerator could be huge. For example, when 5 different types of checkers are available for each of 14 positions in a heat regenerator, the total number of possible arrangements of checkers is 6,103,515,625. It is impractical to completely evaluate the efficiency of each of the 6,103,515,625 arrangements of checkers by 3D CFD simulations on Ansys Fluent. Here, we propose an optimization algorithm by tree search to tackle this optimization problem. This tree search method is motivated by the recent applications of Artificial Intelligence, based on combination of Deep Learning with Monte-Carlo Tree Search, to the incredibly complicated board game Go. Empirical evidence shows that this simple tree search algorithm leads to fast convergence of an optimization search and successfully suggests the optimal arrangement of checkers. This simple tree search method/algorithm may effectively enhance the thermal efficiency of a regenerative combustion system.

MHD Free Convection in an Enclosure Loaded with Nanofluid and Partially Cross-heated

This paper digitally examines natural convection in a porous cavity laden with nanofluids. The influence of Brownian motion and thermophoresis on particle motion through the Buongiorno model is reviewed. The application of the Darcy-Brinkman model makes it possible to simulate the transfer of momentum under the effect of a horizontal magnetic field. Two segments located in the central parts of the left and bottom walls heat the cavity. The right wall is cooled at a constant temperature and the remaining surfaces of the cavity boundary are insulated. A computational code, designed on the principles of the finite volume method in conjunction with the SIMPLE algorithm, is employed to address the fundamental equations governing the system effectively. The numerical results are illustrated by presenting streamlines, isotherms, iso-concentrations and local and mean Nusselt numbers. The dimensionless parameters that change are the Rayleigh number 103 ≤ RT ≤ 106, the Darcy number 10-5 ≤ Da ≤ 10-2 the Hartmann number 0 ≤ Ha ≤ 100, the thermophoresis parameter 0.05 ≤ Nt ≤ 0.5 the Brownian motion parameter 0.05 ≤ Nb ≤ 0.5, and the buoyancy parameter 0.05 ≤ Nr ≤ 0.5. The increase in RT leads to an improvement in the heat transfer rate and an attenuation of the inhomogeneity of the distribution of nanoparticles in the cavity. A reverse tendency is observed by increasing the intensity of the magnetic field. Moreover, the parameters and exhibit a more significant impact on the Sherwood number than on the Nusselt number.

Prototype acoustic positioning system for the Pacific Ocean Neutrino Experiment

We present the design and initial performance characterization of the prototype acoustic positioning system intended for the Pacific Ocean Neutrino Experiment. It comprises novel piezo-acoustic receivers with dedicated filtering- and amplification electronics installed in P-ONE instruments and is complemented by a commercial system comprised of cabled and autonomous acoustic pingers for sub-sea installation manufactured by Sonardyne Ltd. We performed an in-depth characterization of the acoustic receiver electronics and their acoustic sensitivity when integrated into P-ONE pressure housings. These show absolute sensitivities of up to -125 dB re V2/μPa2 in a frequency range of 10–40 kHz. We furthermore conducted a positioning measurement campaign in the ocean by deploying three autonomous acoustic pingers on the seafloor, as well as a cabled acoustic interrogator and a P-ONE prototype module deployed from a ship. Using a simple peak-finding detection algorithm, we observe high accuracy in the tracking of relative ranging times at approximately 230–280 μs at distances of up to 1600 m, which is sufficient for positioning detectors in a cubic-kilometer detector and which can be further improved with more involved detection algorithms. The tracking accuracy is further confirmed by independent ranging of the Sonardyne system and closely follows the ship's drift in the wind measured by GPS. The absolute positioning shows the same tracking accuracy with its absolute precision only limited by the large uncertainties of the deployed pinger positions on the seafloor.

Tight Distance Query Reconstruction for Trees and Graphs Without Long Induced Cycles

ABSTRACTGiven access to the vertex set of a connected graph and an oracle that given two vertices , returns the shortest path distance between and , how many queries are needed to reconstruct ? Firstly, we show that randomized algorithms need to use at least queries in expectation to reconstruct ‐vertex trees of maximum degree . The best previous lower bound (for graphs of bounded maximum degree) was an information‐theoretic lower bound of . Our randomized lower bound is also the first to break through the information‐theoretic barrier for related query models, including distance queries for phylogenetic trees, membership queries for learning partitions, and path queries in directed trees. Secondly, we provide a simple deterministic algorithm to reconstruct trees using distance queries. This proves that our lower bound is optimal up to a multiplicative constant. We extend our algorithm to reconstruct graphs without induced cycles of length at least using queries. Our lower bound is therefore tight for a wide range of tree‐like graphs, such as chordal graphs, permutation graphs, and AT‐free graphs. The previously best randomized algorithm for chordal graphs used queries in expectation, so we improve by a ‐factor for this graph class.

Metabolic Associated Steatotic Liver Disease (MASLD) – Be Aware and Beware

Metabolic associated steatotic liver disease (MASLD) is now the most common liver and metabolic disease with rapidly rising prevalence, being among most common causes of liver transplantation, associated with liver mortality, but even more so and earlier in the course of the disease it is underappreciated independent risk factor for cardiovascular (CV) and all-cause mortality. A large body of clinical evidence suggests that MASLD is a multisystem disease whose adverse effects extend far beyond and before the liver gets seriously affected. It has a complex, independent and bidirectional relationship to all MetS components, chronic kidney disease, and CVD, being causal in one and consequential in another patient and that speaks in favor of including liver health assessment in conventional screening of this at-risk population. Therefore, authors of this editorial call for raising awareness about this condition, write about new nomenclature that better explains what this condition is rather than what it is not, explain how novel simplified positive diagnostic criteria facilitate timely diagnosis and treatment, and offer simple algorithm for evaluation and treatment of liver steatosis in at-risk patients for non hepatologists. Keywords: liver, metabolic syndrome, cardiometabolic risk factors.

A Distance-based Algorithm for Defining Antitrust Markets

We propose a simple algorithm for defining merger-specific geographic antitrust markets based on merging firm proximity. Applying it to over a thousand hypothetical bank mergers, we compare concentration measures in our markets to those defined by the Federal Reserve, which are not merger-specific, finding broad agreement but also offering potential improvements upon current definitions.

Improving classifier decision boundaries and interpretability using nearest neighbors

Neural networks often fail to learn optimal decision boundaries. In this study, we show that these boundaries are typically situated in regions with low training data density, making them highly sensitive to a small number of samples, which, in turn, increases the risk of overfitting. To address this issue, we propose a simple algorithm that performs a weighted average of a sample’s prediction and those of its nearest neighbors (computed in latent space), leading to minor but favorable outcomes across a variety of important performance measures for neural networks. Through diverse evaluations using both self-trained and state-of-the-art pre-trained convolutional neural networks, we show that our framework enhances (i) resistance to label noise, (ii) robustness against adversarial attacks, (iii) classification accuracy, and offers novel approaches for (iv) interpretability. Notably, the interpretability aspect is particularly relevant to the Explainable AI community, as our approach is broadly applicable to any network architecture. Although the improvements may not be large in all four areas, the proposed solution is conceptually simple and requires no modifications to the network architecture, training procedure, or dataset. Unlike prior methods, our approach achieves improvements without introducing trade-offs or necessitating architectural adaptations, while providing actionable insights and theoretical analysis to support its efficacy.

A genetic algorithm framework for algorithmic design through logic circuit optimization

Abstract This paper presents the design of a logic circuit, whose topology is defined through an optimization process carried out by a genetic algorithm with the aim of obtaining an optimal tic-tac-toe player. To achieve this objective, the optimization process is performed by a simple genetic algorithm, and the problem is framed as a maximization problem. Given a logic circuit and a set of Boolean weights on its vertices, the goal is to optimize the performance of the logic circuit as a tic-tac-toe player by optimizing the weights that define its topology, where each set of weights defines a unique configuration and, therefore, a unique circuit. The genetic algorithm treats these sets of weights as individuals in its population. The design of the genetic algorithm involves evaluating various configurations of logic circuits, objective functions, and training models. The presented results showcase an optimal set of these configurations, as well as an analysis of the best player obtained.

A decentralized point of service triple prevention and treatment model for hepatitis B, C and HIV in people who inject drugs in South Africa.

A decentralized point of service triple prevention and treatment model for hepatitis B, C and HIV in people who inject drugs in South Africa.

Territory Planning Algorithms: Graph-Based Sales Coverage Optimization

Sales territory planning is crucial for maximizing sales performance, distributing workload evenly among representatives, and minimizing travel expenses. Manual assignments, rule-based systems, and simple clustering algorithms alike often fall short in terms of scalability, fairness, and adaptability to market dynamics. In this paper, a comprehensive graph-based framework is introduced that treats customers, depots, and travel paths as nodes and edges of a graph structure. The model incorporates multiple key attributes, including customer value, travel distance, and sales representative capacity, to generate geographically coherent, workload-balanced territories. These territories are also aligned with strategic business objectives. The framework supports dynamic adjustments, leveraging advanced feature engineering and preprocessing techniques to adapt to changing sales data and operational conditions. Experimental evaluations demonstrate that graph-based territory planning outperforms traditional approaches in terms of workload equity, the number of unused trips, and overall customer coverage. Additionally, the model's outputs are transparent and interpretable, enabling sales managers to make more informed and confident decisions. Looking forward, the use of real-time data sources, such as live traffic updates and customer activity logs, combined with machine learning approaches, presents an opportunity to enhance responsiveness and territory optimization further. This graph-based approach can also be applied in other domains beyond sales, such as service delivery, field maintenance, and healthcare outreach. The proposed framework offers a practical, scalable, and adaptable solution for modern sales organizations seeking to remain competitive in a complex and highly data-driven environment.

Новый метод восстановления спектрального коэффициента яркости моря на основании спутниковых данных OLCI первого уровня обработки

The paper proposes an alternative method of atmospheric correction using the OLCI satellite data for the Black Sea as an example. Currently, for remote sensing problems, the standard Gordon and Wang atmospheric correction algorithm is used in most cases (GW94). Unfortunately, its operation is often accompanied by the appearance of negative values of the spectral radiance coefficient of the sea (remote sensing reflectance) rs( ) in the shortwave region, which means a sufficient number of physically incorrect values and subsequent incorrect calculation of the concentration of chlorophyll-a and yellow matter. In this paper, a simple algorithm is proposed, built exclusively on analytical formulas, where two procedures of interpolation and extrapolation are conceptually implemented simultaneously, extrapolation - via two channels, interpolation based on the constancy of the color index ratio (CI = rs(412)/ rs(443) = 0.8). Using individual examples of OLCI scanner data, the performance GW94 of the new algorithm was tested for different states of the atmosphere and sea surface by comparing the results with in-kind measurements of the AERONET-OC platforms, with Level-2 data and with the operation of the regional method of additional correction. The new algorithm was tested under the following conditions: clear atmosphere (presence of background aerosol), presence of dust aerosol, cloud boundaries, presence of sun glare, coccolithophore bloom. When analyzing a number of Sentinel 3A/3B satellite images, it was found that the new simple algorithm was, on average, better than the standard one, which means that there is a prospect for its improvement. The advantage of this approach is its universality and the possibility of its implementation for other water areas, if there are patterns in the variability of the "blue" color index.

Understanding the training of infinitely deep and wide ResNets with conditional optimal transport

Abstract We study the convergence of gradient flow for the training of deep neural networks. While residual neural networks (ResNet) are a popular example of very deep architectures, their training constitutes a challenging optimization problem, notably due to the non‐convexity and the non‐coercivity of the objective. Yet, in applications, such tasks are successfully solved by simple optimization algorithms such as gradient descent. To better understand this phenomenon, we focus here on a “mean‐field” model of an infinitely deep and arbitrarily wide ResNet, parameterized by probability measures on the product set of layers and parameters, and with constant marginal on the set of layers. Indeed, in the case of shallow neural networks, mean field models have been proven to benefit from simplified loss landscapes and good theoretical guarantees when trained with gradient flow w.r.t. the Wasserstein metric on the set of probability measures. Motivated by this approach, we propose to train our model with gradient flow w.r.t. the conditional optimal transport (COT) distance: a restriction of the classical Wasserstein distance which enforces our marginal condition. Relying on the theory of gradient flows in metric spaces, we first show the well‐posedness of the gradient flow equation and its consistency with the training of ResNets at finite width. Performing a local Polyak–Łojasiewicz analysis, we then show convergence of the gradient flow for well‐chosen initializations: if the number of features is finite but sufficiently large and the risk is sufficiently small at initialization, the gradient flow converges to a global minimizer. This is the first result of this type for infinitely deep and arbitrarily wide ResNets. In addition, this work is an opportunity to study in more detail the COT metric, particularly its dynamic formulation. Some of our results in this direction might be interesting on their own.

Single-pixel shadowless imaging by a ring detector.

Active single-pixel imaging technology has shown significant advantages in many fields. However, its dependence on a single light source often leads to shadow artifacts, which reduces the imaging quality. Inspired by surgical shadowless lamps, this study proposes an innovative solution based on a ring detector to overcome this limitation. Taking advantage of the fact that the field of view of active single-pixel imaging depends on the projected pattern area, we explored multiple methods to construct a ring detector consisting of eight detectors. Specifically, eight independent detectors capture images with different shadows, and the shadowless effect is achieved by signal and image superposition. A simpler method is to connect eight detectors in series or in parallel to achieve shadowless imaging in a single acquisition. In addition, we combine a large-area ring detector with an active Bayer-Fourier pattern to achieve shadowless color imaging. Through a simple restoration algorithm, we simplify data processing and successfully separate and reconstruct the red, green, and blue color components. We expect that this simple shadowless imaging technology will have a wide range of applications in fields such as medical imaging and industrial inspection.