Personalized marketing is all the rage these days. It is the key to helping users find exactly what they are looking for, quickly and easily. But traditional recommendation systems have their limits. They struggle with complex user behaviors and sparse data. This makes it hard to understand the relationships between different types of interactions and long-term depen...
Water is a superstar in the world of chemistry. It's everywhere, and it's great at breaking down molecules in a process called hydrolysis. This happens all the time in nature and in labs. But predicting how water will react with different substances is tricky. Small changes in a molecule can make a big difference. Plus, the acidity or basicity of the water can change...
Uniform point sets are essential in many fields. They are used in experiments, graphics, and finance. These sets need to have low discrepancy, which means the points are spread out evenly. Recent studies have shown that using Graph Neural Networks and optimization can create point sets with much lower discrepancy than before. However, there is a new approach that co...
Chemicals can be harmful to the environment, but testing each one is a slow process. Scientists have found a way to speed this up using something called graph neural networks (GNNs). These networks can predict how harmful a chemical might be to tiny microbes called nitrifiers. These microbes are important for keeping our water clean. However, GNNs can sometimes give...
Cities around the world have unique road layouts. These layouts can be compared using a new method. This method uses AI to measure how different these layouts are. This is a big deal because traditional methods of comparison have strict limits. These limits make it hard to see the true differences between city road networks. The new method uses something called ...
Imagine trying to guess how busy a city street will be in an hour. It's not just about looking at past traffic data. You also need to think about things like the weather and what's happening nearby. Most methods out there only look at past traffic data for short periods. But what if we could use more info and look further into the future? Enter the MIFPN model, a sm...
Metal complexes, like transition metal complexes and metal-organic frameworks, are super important in fields like catalysis, materials science, and creating tiny molecular devices. This is because they have a unique central metal atom and a complex structure around it. But, predicting how these complexes will behave can be tricky. That's where a new approach called t...
Additive manufacturing (AM) has revolutionized the creation of customized materials with unique properties. It's not just about shapes anymore; now, we can mix different materials in AM to make stuff with even more special features. So far, using machine learning to design these materials has mainly focused on single-material systems. But a new approach is changing t...
Graph neural networks could benefit greatly from specialized hardware. Memristors, especially those made from robust and epitaxial films, can make these networks more efficient and better at handling graph-structured data. These memristors are made from materials like gadolinium-doped hafnium oxide (Gd: HfO2), which are known for their stability and low power consump...
Traffic forecasting on freeways has long relied on deep learning models, especially graph neural networks that blend graph theory with deep learning. These models are great at learning complex time and space patterns. However, traditional graph convolutional networks (GCNs) sometimes struggle with long-range spatial correlations, which can make long-term predictions ...