Related papers: LSTM-based Load Forecasting Robustness Against Noi…

Noisy labels are inevitable in large real-world datasets. In this work, we explore an area understudied by previous works -- how the network's architecture impacts its robustness to noisy labels. We provide a formal framework connecting the…

This work presents a deep-learning approach to estimate atmospheric density profiles for use in planetary entry guidance problems. A long short-term memory (LSTM) neural network is trained to learn the mapping between measurements available…

Short-term industrial enterprises power system forecasting is an important issue for both load control and machine protection. Scientists focus on load forecasting but ignore other valuable electric-meters which should provide guidance of…

Forecasting the actual amount of electricity with respect to the need/demand of the load is always been a challenging task for each power plants based generating stations. Due to uncertain demand of electricity at receiving end of station…

We present a noise-injected version of the Expectation-Maximization (EM) algorithm: the Noisy Expectation Maximization (NEM) algorithm. The NEM algorithm uses noise to speed up the convergence of the EM algorithm. The NEM theorem shows that…

It is well-known that deep neural networks (DNNs) have shown remarkable success in many fields. However, when adding an imperceptible magnitude perturbation on the model input, the model performance might get rapid decrease. To address this…

The query-based black-box attacks have raised serious threats to machine learning models in many real applications. In this work, we study a lightweight defense method, dubbed Random Noise Defense (RND), which adds proper Gaussian noise to…

The key challenge in multispectral radiation thermometry is accurately measuring emissivity. Traditional constrained optimization methods often fail to meet practical requirements in terms of precision, efficiency, and noise resistance.…

In this paper, we address the problem of multichannel speech enhancement in the short-time Fourier transform (STFT) domain. A long short-time memory (LSTM) network takes as input a sequence of STFT coefficients associated with a frequency…

The increasing adoption of measurement units in electrical power distribution grids has enabled the deployment of data-driven and measurement-based control schemes. Such schemes rely on measurement-based estimated models, where the models…

Speech intelligibility can be degraded due to multiple factors, such as noisy environments, technical difficulties or biological conditions. This work is focused on the development of an automatic non-intrusive system for predicting the…

Supervised fine-tuning (SFT) plays a crucial role in adapting large language models (LLMs) to specific domains or tasks. However, as demonstrated by empirical experiments, the collected data inevitably contains noise in practical…

Neural network (NN) emulators of the global 21 cm signal need emulation error much less than the observational noise in order to be used to perform unbiased Bayesian parameter inference. To this end, we introduce $\texttt{21cmLSTM}$ -- a…

Large language models (LLMs) often produce inaccurate or misleading content-hallucinations. To address this challenge, we introduce Noise-Augmented Fine-Tuning (NoiseFiT), a novel framework that leverages adaptive noise injection based on…

In this paper we study how the choice of loss functions of non-convex optimization problems affects their robustness and optimization landscape, through the study of noisy matrix sensing. In traditional regression tasks, mean squared error…

Estimating the number of signals embedded in noise is a fundamental problem in array signal processing. The classic RMT estimator based on random matrix theory (RMT) tends to under-estimate the number of signals as it does not consider the…

The bias-compensated set-membership normalised LMS (BCSMNLMS) algorithm is proposed based on the concept of set-membership filtering, which incorporates the bias-compensation technique to mitigate the negative effect of noisy inputs.…

We explore the robustness of recurrent neural networks when the computations within the network are noisy. One of the motivations for looking into this problem is to reduce the high power cost of conventional computing of neural network…

This paper focuses on the application and optimization of LSTM model in financial risk prediction. The study starts with an overview of the architecture and algorithm foundation of LSTM, and then details the model training process and…

While Large Language Models (LLMs) have shown impressive capabilities in math problem-solving tasks, their robustness to noisy inputs is not well-studied. We propose ArithmAttack to examine how robust the LLMs are when they encounter noisy…