New research reveals that AI-powered predictive policing systems, even with attempts at data correction, can worsen existing biases and lead to significantly unequal outcomes.
Researchers have developed a novel method for creating realistic, high-fidelity anomalous images without the need for extensive training data.
Distinguishing between genuine equipment failures and normal operational changes is critical for reliable industrial automation, and new research offers a way to help systems – and human operators – tell the difference.
![A feedforward neural network predicts time series data by leveraging lagged inputs-[latex]x\_{t-1},\dots,x\_{t-p}[/latex]-and an inverse transformation-[latex]t^{-1}(\cdot)[/latex]-maps network weights to autoregressive coefficients, effectively ensuring the stationarity of the predicted series and highlighting the model’s capacity to navigate the precarious balance between complexity and stability.](https://arxiv.org/html/2603.19041v1/NN.png)
A new approach combines the power of neural networks with the established principles of autoregressive modeling for faster, more reliable time series forecasting.
![The analysis reveals that each model contributes a distinct share to the overall prevalence of anticipatory outliers-those instances flagged as problematic before they manifest-within the broader landscape of all detected outliers [latex]\mathcal{TOA}[/latex] and general outliers [latex]\mathcal{TO}, \mathcal{O}[/latex].](https://arxiv.org/html/2603.18358v1/images/seed_to_topic_barplot-v2.png)
New research details how seemingly anomalous data points in news streams can actually foreshadow the development of dominant topics over time.
A new autoregressive approach to generative classification is challenging the dominance of diffusion models, delivering superior performance and enhanced robustness.
New research reveals that while firms invest heavily in artificial intelligence, this often correlates with a rise in consumer complaints and negative emotional responses.
A new framework introduces a way to quantify and improve the diversity of images created by generative adversarial networks by explicitly acknowledging what the AI doesn’t know.
Researchers have developed a reinforcement learning-based method to efficiently identify similar subgraphs within larger network structures, offering improvements over traditional search algorithms.
This review examines how advanced artificial intelligence models are being used to automatically identify crucial design conversations within the often-noisy landscape of modern software development.