long short-term memory
PulseAugur coverage of long short-term memory — every cluster mentioning long short-term memory across labs, papers, and developer communities, ranked by signal.
- uses graph attention network 70%
- used by Shap 70%
- competes with Arima 70%
- developed by xLSTM: Extended Long Short-Term Memory 70%
- competes with xLSTM: Extended Long Short-Term Memory 70%
- used by gate 70%
- competes with Gated Recurrent Unit 70%
- competes with support vector machine 60%
- other Gated Recurrent Unit 50%
- competes with graph attention network 50%
- 2026-05-14 research_milestone A hybrid LSTM model achieved the lowest final displacement error in dynamic movement forecasting. 来源
8 天有情绪数据
-
New X-TRACK model uses xLSTM and physics for realistic vehicle trajectory prediction
Researchers have developed X-TRACK, a novel trajectory prediction model for autonomous driving that leverages the extended Long Short-Term Memory (xLSTM) architecture. This new model explicitly incorporates vehicle moti…
-
Hybrid KAN-XGBoost model improves electricity price forecasting
Researchers have developed a new hybrid framework for forecasting electricity prices in Australia's National Electricity Market (NEM). This approach combines Kolmogorov-Arnold Networks (KAN) with XGBoost to better captu…
-
RoBERTa leads sentiment analysis with 93% accuracy in new study
This paper explores sentiment classification using various machine learning models, including traditional methods like Naive Bayes and SVM, alongside transformer-based models such as RoBERTa and DistilBERT. The study ev…
-
AI models explore music generation using piano and MIDI data
Researchers are exploring AI models for music generation, with one project focusing on creating generative instruments using large piano models trained on performance data. Another initiative details building an AI mode…
-
Weight decay controls transformer training regimes, new diagnostics revealed
Researchers have identified weight decay as a key parameter controlling the training regimes of transformers on modular arithmetic tasks. They introduced two new, low-cost online diagnostics—mean pairwise attention-head…
-
CogScale benchmark accelerates AI sequence processing evaluation
Researchers have introduced CogScale, a new benchmark designed to efficiently evaluate the sequential processing capabilities of AI architectures. This benchmark comprises 14 scalable synthetic tasks that allow for rapi…
-
GRU: A simpler, faster successor to LSTM for sequence modeling
The Gated Recurrent Unit (GRU) was developed in 2014 as a simpler alternative to the Long Short-Term Memory (LSTM) network. While LSTM uses separate cell and hidden states with three gates, GRU consolidates these into a…
-
Shipping logistics boosted by new retrieval-enhanced Transformer model
Researchers have developed a novel deep learning framework called CCRE to improve multi-step port-of-call sequence prediction in global shipping logistics. This framework utilizes a retrieval-enhanced historical encoder…
-
TinyDL model detects logic anomalies in industrial water treatment
Researchers have developed a new lightweight anomaly detection model called Ti-iLSTM, designed for resource-constrained industrial control systems. This Tiny Deep Learning (TinyDL) approach optimizes Long Short-Term Mem…
-
Hybrid LSTM model leads in NBA player movement forecasting
Researchers have explored various neural network architectures for dynamic movement forecasting, particularly in the context of NBA player trajectories. Traditional methods like Kalman filters struggle with the non-line…
-
Diffusion model and LSTM optimize radiotherapy plans
Researchers have developed a novel diffusion model and LSTM-based approach for optimizing radiotherapy plans, specifically for Volumetric Modulated Arc Therapy (VMAT). This method aims to significantly reduce the planni…
-
Time series forecasting explained with Zillow's $9B lesson
This article explains time series forecasting, a crucial but often complex aspect of data analysis. It uses the example of Zillow's costly failure in iBuying to illustrate the dangers of models that don't account for ch…
-
LSTM networks overcome RNN memory limitations with gating mechanisms
The Long Short-Term Memory (LSTM) network was developed to address the limitations of traditional Recurrent Neural Networks (RNNs) in handling sequential data. Vanilla RNNs struggle with remembering information over lon…
-
CatNet paper introduces SHAP for feature importance in LSTM FDR control
Researchers have introduced CatNet, a novel algorithm designed to control the False Discovery Rate (FDR) and identify significant features within Long Short-Term Memory (LSTM) networks. This method utilizes the derivati…
-
Neural-Actuarial Longevity Forecasting: Anchoring LSTMs for Explainable Risk Management
Researchers have developed a new neural-actuarial framework called Hybrid-Lift to improve longevity forecasting. This approach combines Hierarchical LSTM networks with a Mean-Bias Correction anchoring mechanism to addre…
-
Towards AI explains fundamentals of time series analysis before model application
This guide explains the fundamental differences between standard machine learning data and time series data, emphasizing that the order of observations is crucial in time series. It details various types of time series …
-
Geometric Evolution Graph Convolutional Networks enhance graph representation learning
Researchers have developed a new framework called the Geometric Evolution Graph Convolutional Network (GEGCN) to improve graph representation learning. This novel approach utilizes a Long Short-Term Memory (LSTM) networ…
-
Hybrid quantum-classical model enhances financial volatility forecasting
Researchers have developed a hybrid quantum-classical framework for financial volatility forecasting, integrating a Long Short-Term Memory (LSTM) network with a Quantum Circuit Born Machine (QCBM). The LSTM extracts tem…
-
Auto-encoders and PCA accelerate phase-field simulations with 80% accuracy
Researchers have developed a data-driven framework using auto-encoder neural networks and principal component analysis to significantly reduce the dimensionality of simulated microstructural images, achieving a reductio…
-
AI governance framework integrates US banking regulations for fraud detection
Researchers have developed a new framework to help U.S. financial institutions navigate the complex regulatory landscape for AI-driven fraud detection. This framework, called RGF-AFFD, integrates requirements from four …