Filtreler

2011 - 201912020 - 20231
Filtreleri Sıfırla

Ayarlar

search.filters.applied.f.publicationcategory: Konferans Öğesi - Uluslararası - Kurum Öğretim Elemanı × Konu: Heuristic methods ×

Arama Sonuçları

Listeleniyor 1 - 2 / 2
  • Küçük Resim
    Yayın
    Mixture of Gaussian models and bayes error under differential privacy
    (2011) Xi, Bowei; Kantarcıoğlu, Murat; İnan, Ali
    Gaussian mixture models are an important tool in Bayesian decision theory. In this study, we focus on building such models over statistical database protected under differential privacy. Our approach involves querying necessary statistics from a database and building a Bayesian classifier over the noise added responses generated according to differential privacy. We formally analyze the sensitivity of our query set. Since there are multiple methods to query a statistic, either directly or indirectly, we analyze the sensitivities for different querying methods. Furthermore we establish theoretical bounds for the Bayes error for the univariate (one dimensional) case. We study the Bayes error for the multivariate (high dimensional) case in experiments with both simulated data and real life data. We discover that adding Laplace noise to a statistic under certain constraint is problematic. For example variance-covariance matrix is no longer positive definite after noise addition. We propose a heuristic method to fix the noise added variance-covariance matrix.
  • Küçük Resim
    Yayın
    Optimizing indoor localization accuracy with neural network performance metrics and software-defined IEEE 802.11az Wi-Fi set-up
    (IEEE, 2023-10-28) Kouhalvandi, Lida; Aygün, Sercan; Matekovits, Ladislau; Miramirkhani, Farshad
    Accurately classifying regions based on Wi-Fi signals can be a difficult task, especially when considering different frequency values. In this study, we aimed to improve the accuracy of indoor localization by developing a novel approach that does not rely on pre-trained models. To achieve this, fingerprints from the IEEE 802.11az standard were randomly selected, and the data samples were trained using parameterized station characteristics and neural network hyperparameters. The impact of each parameter on the localization accuracy was measured, and performance monitoring metrics such as F1-Measure and confusion matrix-based metrics were evaluated. Furthermore, the Thompson sampling (TS) algorithm was employed to determine the optimal parameters, which helped to achieve the best possible accuracy. The proposed approach demonstrated improved accuracy in region localization compared to conventional heuristic approaches which typically yield an accuracy range of 65% to 77%. The proposed approach achieved up to 80% accuracy in region localization and could be a promising solution for indoor localization in various settings.