The integration of hardware accelerators has significantly advanced the capabilities of modern recommendation systems, enabling the exploration of complex ranking paradigms previously deemed impractical. However, the GPU-based computational costs present substantial challenges. In this paper, we demonstrate our development of an efficiency-driven approach to explore these paradigms, moving beyond traditional reliance on native PyTorch modules. We address the specific challenges posed by ranking models’ dependence on categorical features, which vary in length and complicate GPU utilization. We introduce Jagged Feature Interaction Kernels, a novel method designed to extract fine-grained insights from long categorical features through efficient handling of dynamically sized tensors. We further enhance the performance of attention mechanisms by integrating Jagged tensors with Flash Attention. Our novel Jagged Flash Attention achieves up to 9 × speedup and 22 × memory reduction compared to dense attention. Notably, it also outperforms dense flash attention, with up to 3 × speedup and 53% more memory efficiency. In production models, we observe 10% QPS improvement and 18% memory savings, enabling us to scale our recommendation systems with longer features and more complex architectures.

Projection and ranking are frequently used analysis techniques in multi-attribute data exploration. Both families of techniques help analysts with tasks such as identifying similarities between observations and determining ordered subgroups, and have shown good performances in multi-attribute data exploration. However, they often exhibit problems such as distorted projection layouts, obscure semantic interpretations, and non-intuitive effects produced by selecting a subset of (weighted) attributes. Moreover, few studies have attempted to combine projection and ranking into the same exploration space to complement each other's strengths and weaknesses. For this reason, we propose RankAxis, a visual analytics system that systematically combines projection and ranking to facilitate the mutual interpretation of these two techniques and jointly support multi-attribute data exploration. A real-world case study, expert feedback, and a user study demonstrate the efficacy of RankAxis.

Accelerating finite automata processing is critical for advancing real-time analytic in pattern matching, data mining, bioinformatics, intrusion detection, and machine learning. Recent in-memory automata accelerators leveraging SRAMs and DRAMs have shown exciting improvements over conventional digital designs. However, the bit-vector representation of state transitions used by all state-of-the-art (SOTA) designs is only optimal in processing worst-case completely random patterns, while a significant amount of memory and energy is wasted in running most real-world benchmarks.We present CAMA, a Content-Addressable Memory (CAM) enabled Automata accelerator for processing homogeneous non-deterministic finite automata (NFA). A radically different state representation scheme, along with co-designed novel circuits and data encoding schemes, greatly reduces energy, memory, and chip area for most realistic NFAs. CAMA is holistically optimized with the following major contributions: (1) a 16 × 256 8-transistor (8T) CAM array for state matching, replacing the 256 × 256 6T SRAM array or two 16×256 6T SRAM banks in state-of-the-art (SOTA) designs; (2) a novel encoding scheme that enables content searching within 8T SRAMs and adapts to different applications; (3) a reconfigurable and scalable architecture that improves efficiency on all tested benchmarks, without losing support for any NFA that’s compatible with SOTA designs; (4) an optimization framework that automates the choice of encoding schemes and maps a given NFA to the proposed hardware.Two versions of CAMA, one optimized for energy (CAMA-E) and the other for throughput (CAMA-T), are comprehensively evaluated in a 28nm CMOS process, and across 21 real-world and synthetic benchmarks. CAMA-E achieves 2.1×, 2.8 ×, and 2.04× lower energy than CA, 2-stride Impala, and eAP. CAMA-T shows 2.68×, 3.87× and 2.62 × higher average compute density than 2-stride Impala, CA, and eAP. Both versions reduce the chip area required for the largest tested benchmark by 2.48× over CA, 1.91× over 2-stride Impala, and 1.78× over eAP.

Lattice-based cryptography (LBC), exploiting learning with error (LWE) problems, is a promising candidate for postquantum cryptography. The number theoretic transform (NTT) is the latency- and energy-dominant process in the computation of LWE problems. This article presents a compact and efficient in-MEmory NTT accelerator, named MeNTT, which explores an optimized computation in and near a 6T SRAM array. Specifically designed peripherals enable fast and efficient modular operations. Moreover, a novel mapping strategy reduces the data flow between the NTT stages into a unique pattern, which greatly simplifies the routing among processing units (i.e., the SRAM column in this work), reducing the energy and area overheads. The accelerator achieves significant latency and energy reductions over prior arts.

In urban traffic networks, abundant traffic data is generated everyday. Therefore, data-driven control approaches for urban traffic control without utilizing mathematical models might be a promising research direction. In this paper, a hierarchical perimeter control strategy for urban traffic networks based on model-free adaptive control (MFAC) scheme is proposed. At the outer level, the signal settings in the periphery of the congested region are optimized to meter the number of vehicles entering the congested region; while at the inner level, traffic flow distribution within the congested region is adjusted to improve the traffic efficiency. MFAC scheme is applied in both outer level and inner level to derive the detailed signal settings in the congested region and in the periphery of the congested region. Joint simulations using VISSIM and MATLAB verify the feasibility and effectiveness of the proposed method.

This paper presents an automatic text document categorization system, HPHR. HPHR contains high precision, high recall and noise-filtered text categorization models. The text categorization models are generated through a suite of machine learning algorithms, a fast clustering algorithm that efficiently and effectively group documents into subcategories, and a text category generation algorithm that automatically generates text subcategories that represent high precision, high recall and noise-filtered text categorization models from a given set of training documents. The HPHR system was evaluated on documents drawn from two different applications, vehicle fault diagnostic documents, which are in a form of unstructured and verbatim text descriptions, and Reuters corpus. The performance of the proposed system, HPHR, on both document collections showed superiority over the systems commonly used in text document categorization.

Prediction of the traffic information such as flow, density, speed, and travel time is important for traffic control systems, optimizing vehicle operations, and the individual driver. Prediction of future traffic information is a challenging problem due to many dynamic contributing factors. In this paper, various methodologies for traffic information prediction are investigated. We present a speed prediction algorithm, NNTM-SP (Neural Network Traffic Modeling-Speed Prediction) that trained with the historical traffic data and is capable of predicting the vehicle speed profile with the current traffic information. Experimental results show that the proposed algorithm gave good prediction results on real traffic data and the predicted speed profile shows that NNTM-SP correctly predicts the dynamic traffic changes.

This paper presents the methodologies developed for solving IJCNN 2011's Ford Challenge II problem, where the driver's alertness is to be detected employing physiological, environmental and vehicular data acquired during driving. The solution is based on a thorough four-fold framework consisting of temporal processing, feature creation and extraction, and the training and ensemble of several learning systems, such as neural networks, random forest, support vector machine, trained from diverse features. The selection of input features to a learning machine has always been critique on signal classification. In our approach, the employment of Bayesian network filtered out a set of features and has been proved by the ensemble to be effective. The ensemble technique enhanced the performance of individual systems dramatically. The performance acquired on 30% of the test samples reached an accuracy of 78.34%. These results are significant for a real-world vehicular problem and we are quite confident this solution will become one of the top ones on the competition test data.