Chien, Steve A., Rabideau, Gregg R. · NASA STI Repository (National Aeronautics and Space Administration)
AVA v2 software selects goals for execution from a set of goals that oversubscribe shared resources. The term goal refers to a science or engineering request to execute a possibly complex command sequence, such as image targets or ground-station downlinks. Developed as an extension to the Virtual Machine Language (VML) execution system, the software enables onboard and remote goal triggering through the use of an embedded, dynamic goal set that can oversubscribe resources. From the set of conflicting goals, a subset must be chosen that maximizes a given quality metric, which in this case is strict priority selection. A goal can never be pre-empted by a lower priority goal, and high-level goals can be added, removed, or updated at any time, and the "best" goals will be selected for execution. The software addresses the issue of re-planning that must be performed in a short time frame by the embedded system where computational resources are constrained. In particular, the algorithm addresses problems with well-defined goal requests without temporal flexibility that oversubscribes available resources. By using a fast, incremental algorithm, goal selection can be postponed in a "just-in-time" fashion allowing requests to be changed or added at the last minute. Thereby enabling shorter response times and greater autonomy for the system under control.
Cited by 0Open access
Shivam Gupta, Shweta Jain, Narayanan C. Krishnan et al. · Frontiers in artificial intelligence and applications · 2024
Clustering is a widely used unsupervised learning tool with applications in numerous real-world problems. Traditional clustering methods can result in highly skewed clusters where one cluster is notably larger than others, rendering them unsuitable for scenarios such as logistics and routing. In response, capacitated clustering approaches have emerged over the past decade. These approaches limit the number of data points each cluster can accommodate, thus resulting in more uniform cluster formations. In an online version of capacitated clustering, the algorithm must make an irrevocable decision for each incoming data point, determining whether to establish it as a new center or allocate it to existing centers. The goal is to minimize the count of opened centers while adhering to capacity constraints and achieving a satisfactory approximation of the clustering cost compared to the optimal solution. Although exploring online capacitated clustering remains uncharted, we are the first to propose a probabilistic Capacitated Online Clustering Algorithm (called COCA) for h-dimensional euclidean spaces. We theoretically bound the number of centers opened and provide constant cost approximation guarantees. Additionally, we conduct rigorous experiments to validate the computational efficacy of the proposed approaches.
Cited by 0Open access
Kazem Meidani, Seyedali Mirjalili, Amir Barati Farimani · Expert Systems with Applications · 2022
The performance of optimization algorithms significantly depends on the landscape of the problems. It is known that there is no single algorithm that outperforms others on problems with different fitness landscapes. One of the issues in metaheuristic algorithms is keeping the balance between exploration and exploitation. The features extracted from analysis of fitness landscapes can be used to select the suitable algorithm for the given problem. However, these features are usually expensive and extracted prior to the optimization process which leads to a single algorithm to be selected. In this work, we propose an intelligent switch mechanism that enjoys an efficient non-convex ratio (ENCR) feature extracted online during the optimization to switch between two choices of algorithms, each favoring a type of landscape in terms of modality. For this work, two case studies including a pair of Harris hawks optimizer (HHO) and differential evolution (DE) and another pair of multiverse optimizer (MVO) and moth-flame optimizer (MFO) are selected among several algorithms to evaluate the performance of this framework. The proposed one-way and two-way switch algorithms take advantage of the merits of the two base algorithms to reach better final solutions and higher convergence rates in the majority of case studies. The overall comparison and ranking of the algorithms, including a random switch baseline, demonstrates the superiority of the intelligent switch mechanisms over the baselines.
Cited by 18Open access
Shivam Gupta, Shweta Jain, Narayanan C Krishnan et al. · Zenodo (CERN European Organization for Nuclear Research) · 2024
Appendix for "Capacitated Online Clustering Algorithm" paper accepted in European Conference in Artificial Intelligence 2024. The abstract of the paper is: Clustering is a widely used unsupervised learning tool with applications in numerous real-world problems. Traditional clustering methods can result in highly skewed clusters where one cluster is notably larger than others, rendering them unsuitable for scenarios such as logistics and routing. In response, capacitated clustering approaches have emerged over the past decade. These approaches limit the number of data points each cluster can accommodate, thus resulting in more uniform cluster formations. In an online version of capacitated clustering, the algorithm must make an irrevocable decision for each incoming data point, determining whether to establish it as a new center or allocate it to existing centers. The goal is to minimize the count of opened centers while adhering to capacity constraints and achieving a satisfactory approximation of the clustering cost compared to the optimal solution. Although exploring online capacitated clustering remains uncharted, we are the first to propose a probabilistic Capacitated Online Clustering Algorithm (called COCA) for h-dimensional euclidean spaces. We theoretically bound the number of centers opened and provide constant cost approximation guarantees. Additionally, we conduct rigorous experiments to validate the computational efficacy of the proposed approaches.
Cited by 0Open access
Nicolas Le Roux, Pierre-Antoine Manzagol, Yoshua Bengio · 2007
Guided by the goal of obtaining an optimization algorithm that is both fast and yields good generalization, we study the descent direction maximizing the decrease in generalization error or the probability of not increasing generalization error. The surprising result is that from both the Bayesian and frequentist perspectives this can yield the natural gradient direction. Although that direction can be very expensive to compute we develop an efficient, general, online approximation to the natural gradient descent which is suited to large scale problems. We report experimental results showing much faster convergence in computation time and in number of iterations with TONGA (Topmoumoute Online natural Gradient Algorithm) than with stochastic gradient descent, even on very large datasets.
Cited by 125
LV Patil. · International Journal of Advanced Research · 2017
The first step in problem solving is to develop a foolproof algorithm for the same. This algorithm encompasses of the central logic that will be used to solve the problem. The logical steps that are followed
Cited by 1Open access
Nicolas Le Roux, Yoshua Bengio, Pierre-antoine Manzagol · 2008
Guided by the goal of obtaining an optimization algorithm that is both fast and yielding good generalization, we study the descent direction maximizing the decrease in generalization error or the probability of not increasing generalization error. The surprising result is that from both the Bayesian and frequentist perspectives this can yield the natural gradient direction. Although that direction can be very expensive to compute we develop an efficient, general, online approximation to the natural gradient descent which is suited to large scale problems. We report experimental results showing much faster convergence in computation time and in number of iterations with TONGA (Topmoumoute Online natural Gradient Algorithm) than with stochastic gradient descent, even on very large datasets.
Cited by 1
Hien D. Nguyen, Florence Forbes, Gersende Fort et al. · Studies in classification, data analysis, and knowledge organization · 2023
Cited by 2Open access
Fehim Köylü · 2017
Data mining is the process of extracting a meaningful information from raw data. Classical data mining algorithms can be used to extract an offline static model for classification problems which has a collected dataset. Unfortunately, Offline algorithms cannot give a solution for nowadays' technologies with streaming data. Streaming algorithms are proposed to deal with data-streams for online learning in the literature. In this paper, a new online classification rule learning algorithm based on a swarm based optimization algorithm, Artificial Bee Colony algorithm, is proposed. Experimental results obtained in this study conducted that the proposed algorithm can be used for data-streams.
Cited by 3
Shen Y Uong Wong, Keem Siah Y Ap, Hwa Jen Y Ap · 2014
This paper presents a Hybrid Fuzzy ARTMAP (FAM) and Online Extreme learning machine (OELM), hereafter denoted as FAM-OELM, which enables online learning to start from the first trained data samples without having to set up an initialization phase which requires a chunk of data samples to be ready prior to training. The idea of developing FAM-OELM is motivated by the ELM concept proposed by Huang et al., for being an efficient learning algorithm that provides better generalization performance at a much faster learning speed. However, different from the batch learning ELM and its variant called the online sequential extreme learning machine which still requires an initial offline training phase before it can turn into online training, the proposed FAM-OELM showcases a framework that enable online learning to commence right from the first data sample. Here, classification can be conducted at any time during the training phase. Such appealing feature of the proposed algorithm has strictly fulfilled the criteria of being truly sequential, while many of the existing algorithms are not. In addition, FAM-OELM automatically grows hidden neuron such that the network can accommodate new information without over fitting and compromising on the knowledge learnt earlier. The simulation results reveal the efficacy and validity of FAM-OELM when it is applied to a real world application and various benchmark problems.
Cited by 0
Bui Huy Khoi · Advances in marketing, customer relationship management, and e-services book series · 2021
In recent years, the internet market in Vietnam has developed strongly, leading to many forms of electronic business being born and with rapid growth. Online shopping is seen as the solution to cost-effective, fast, and convenient shopping. This chapter was based on concepts and theoretical grounds related to behavior intention to accept and apply new technology to form the proposed research model. It consists of three independent factors—(1) utility perception, (2) trust, (3) social influence—and one dependent factor, online purchasing intention. This study reveals the optimal choice by the AIC algorithm.
Cited by 1
Rutuja Powar, Prashant Gawande, Sanjay Dambhare · 2022
The modern power system is bigger, more linked, integrated with renewable energy sources and displaying complicated nonlinear dynamic behaviour. Coherent behaviour involves dividing the system’s machines into groups that exhibit the same behaviour. The coherency of the generators inside the islands formed following a disturbance depends on their stability, demonstrating the significance of accurately identifying coherent generators. This paper aims at developing an online coherency detection algorithm. The development of coherency can significantly reduce the computations required for performing stability studies and produces accurate results. Coherent group formation is influenced by the type and location of the disruption. Model reduction approaches are widely employed with large-scale complicated power systems to improve the performance of simulations. The Dynamic Time Warping algorithm was tested, and results have been shown. An assessment of DTW has been seen. This algorithm has been implemented on the IEEE 68-Bus, 16-Machine, 5-Area System for testing its reliability.
Cited by 1
Yi Li, Philip M. Long · Machine Learning · 2002
Cited by 229
Liang Yu, Yang Yu, Wenjie Wei et al. · ArXiv.org · 2025
Binary Spiking Neural Networks (BSNNs) offer promising efficiency advantages for resource-constrained computing. However, their training algorithms often require substantial memory overhead due to latent weights storage and temporal processing requirements. To address this issue, we propose Binary Spiking Online (BSO) optimization algorithm, a novel online training algorithm that significantly reduces training memory. BSO directly updates weights through flip signals under the online training framework. These signals are triggered when the product of gradient momentum and weights exceeds a threshold, eliminating the need for latent weights during training. To enhance performance, we propose T-BSO, a temporal-aware variant that leverages the inherent temporal dynamics of BSNNs by capturing gradient information across time steps for adaptive threshold adjustment. Theoretical analysis establishes convergence guarantees for both BSO and T-BSO, with formal regret bounds characterizing their convergence rates. Extensive experiments demonstrate that both BSO and T-BSO achieve superior optimization performance compared to existing training methods for BSNNs. The codes are available at https://github.com/hamings1/BSO.
Cited by 0Open access
Sahil Verma, Akash Sinha, Prabhat Kumar et al. · 2020
Recent advances in the technology has led to an exponential growth in the e-commerce business. Users today prefer Online Shopping to legacy shopping methods in order to save their effort and time. With the aim of attracting a greater number of customers, the online shopping portals offer products at cheaper rate than offline retailers. The increase in the number of online shopping portals are making it more difficult for the users to purchase a list of products at minimum cost. This can be attributed to the fact that different shopping portals offer different prices for the same product. This problem of obtaining a sequence of products to be procured from different shopping portals with the aim of minimizing the total cost has been termed as the Internet Shopping Optimization Problem. Owing to the NP-hardness of the problem, this paper proposes the application of Genetic Algorithm to solve the Internet Shopping Optimization Problem. Results obtained clearly indicates the efficiency of the proposed work in achieving the optimal solutions in polynomial time.
Cited by 6
Yi Li, Philip M. Long · 1999
We describe a new incremental algorithm for training linear threshold functions: the Relaxed Online Maximum Margin Algorithm, or ROMMA. ROMMA can be viewed as an approximation to the algorithm that repeatedly chooses the hyperplane that classifies previously seen examples correctly with the maximum margin. It is known that such a maximum-margin hypothesis can be computed by minimizing the length of the weight vector subject to a number of linear constraints. ROMMA works by maintaining a relatively simple relaxation of these constraints that can be efficiently updated. We prove a mistake bound for ROMMA that is the same as that proved for the perceptron algorithm. Our analysis implies that the more computationally intensive maximum-margin algorithm also satisfies this mistake bound; this is the first worst-case performance guarantee for this algorithm. We describe some experiments using ROMMA and a variant that updates its hypothesis more aggressively as batch algorithms t...
Cited by 0
Zhenyu SHOU, Jiangbo QIAN, Yihong DONG et al. · SHILAP Revista de lepidopterología · 2020
Many unsupervised learning to hash algorithm needs to load all data to memory in the training phase,which will occupy a large memory space and cannot be applied to streaming data.An unsupervised online learning to hash algorithm called evolutionary forest hash (EFH) was proposed.In a large-scale data retrieval scenario,the improved evolution tree can be used to learn the spatial topology of the data.A path coding strategy was proposed to map leaf nodes to similarity-preserved binary code.To further improve the querying performance,ensemble learning was combined,and an online evolving forest hashing method was proposed based on the evolving trees.Finally,the feasibility of this method was proved by experiments on two widely used data sets.
Cited by 0Open access
Sirui Li, Kaiqi Zhang, Hongbo Yin et al. · Lecture notes in computer science · 2021
Cited by 4
Gang Wang, Nini Rao, Zhilin Zhang et al. · Lecture notes in computer science · 2006
Cited by 5
Olivier Cappé · 2009 IEEE/SP 15th Workshop on Statistical Signal Processing · 2009
Online (or recursive) estimation of fixed model parameters in general state-space models is a crucial but often difficult task. This paper is about likelihood-based point estimation, showing that an online EM (Expectation-Maximization) algorithm recently proposed for discrete hidden Markov models can be extended to more general settings, including non-linear non-Gaussian state-space models that necessitate the use of sequential Monte Carlo filtering approximations. The performance of the proposed online sequential Monte Carlo EM algorithm is illustrated on numerical examples.
Cited by 55