Numerical Analysis and Scientific Computing Commons^™

T-Sciq: Teaching Multimodal Chain-Of-Thought Reasoning Via Large Language Model Signals For Science Question Answering, Lei Wang, Yi Hu, Jiabang He, Xing Xu, Ning Liu, Hui Liu, Heng Tao Shen

Research Collection School Of Computing and Information Systems

Large Language Models (LLMs) have recently demonstrated exceptional performance in various Natural Language Processing (NLP) tasks. They have also shown the ability to perform chain-of-thought (CoT) reasoning to solve complex problems. Recent studies have explored CoT reasoning in complex multimodal scenarios, such as the science question answering task, by fine-tuning multimodal models with high-quality human-annotated CoT rationales. However, collecting high-quality COT rationales is usually time-consuming and costly. Besides, the annotated rationales are hardly accurate due to the external essential information missed. To address these issues, we propose a novel method termed T-SciQ that aims at teaching science question answering with …

Weakly Supervised Attention-Based Recognition Under Spectral, Turbulence, And Resource Variations, Kshitij Naresh Nikhal

Dissertations and Doctoral Documents, University of Nebraska-Lincoln, 2023–

While supervised optimization paradigms are ubiquitous across diverse recognition systems, the risk of over-fitting and increasing bias have limited their applicability.

This dissertation focuses on unsupervised learning—learning without precisely curated data—and argues that unsupervised learning methods can enable both discriminability and generalizability. Through the use of attention-based machine learning and advanced clustering, unsupervised methods are able to focus on fine-grained information in images without any explicit supervision. The dissertation introduces a domain-bridging framework for tasks like cross-spectrum matching and long-range recognition, utilizing intra-domain clustering and inter-domain matching to generate pseudo-labels. Additionally, a hash-based network is proposed to accelerate the search …

Improvement Of Flow Coefficient Estimation With Limited Well Test Data For Real-Time Condition Analytics Of Choke Valve, T. Visawameteekul, Tin Seong Kam, P. Thamvechvitee

Research Collection School Of Computing and Information Systems

The study discusses a method for monitoring the internal condition of choke valves to predict sand erosion using flow coefficient (Cv). The method calculates the Cv of the choke valve by utilizing eight parameters and compares it to the newly manufactured value to generate warnings. However, the availability of spot-check well test data can significantly impact the model's efficiency if tests are performed infrequently. To address this issue, the Extended Cv monitoring method is proposed in this paper. The main purpose of this study is to develop a model for estimating Cv value in the absence of well test data. …

Non-Binary Evaluation Of Next-Basket Food Recommendation, Yue Liu, Palakorn Achananuparp, Ee-Peng Lim

Research Collection School Of Computing and Information Systems

Next-basket recommendation (NBR) is a recommendation task that predicts a basket or a set of items a user is likely to adopt next based on his/her history of basket adoption sequences. It enables a wide range of novel applications and services from predicting next basket of items for grocery shopping to recommending food items a user is likely to consume together in the next meal. Even though much progress has been made in the algorithmic NBR research over the years, little research has been done to broaden knowledge about the evaluation of NBR methods, which is largely based on the …

Gsword: Gpu-Accelerated Sampling For Subgraph Counting, Chang Ye, Yuchen Li, Shixuan Sun, Wentian Guo

Research Collection School Of Computing and Information Systems

Subgraph counting is a fundamental component for many downstream applications such as graph representation learning and query optimization. Since obtaining the exact count is often intractable, there have been a plethora of approximation methods on graph sampling techniques. Nonetheless, the state-of-the-art sampling methods still require massive samples to produce accurate approximations on large data graphs. We propose gSWORD, a GPU framework that leverages the massive parallelism of GPUs to accelerate iterative sampling algorithms for subgraph counting. Despite the embarrassingly parallel nature of the samples, there are unique challenges in accelerating subgraph counting due to its irregular computation logic. To address …

Understanding The Impact Of Trade Policy Effect Uncertainty On Firm-Level Innovation Investment: A Deep Learning Approach, Daniel Chang, Nan Hu, Peng Liang, Morgan Swink

Research Collection School Of Computing and Information Systems

Integrating the real options perspective and resource dependence theory, this study examines how firms adjust their innovation investments to trade policy effect uncertainty (TPEU), a less studied type of firm specific, perceived environmental uncertainty in which managers have difficulty predicting how potential policy changes will affect business operations. To develop a text-based, context-dependent, time-varying measure of firm-level perceived TPEU, we apply Bidirectional Encoder Representations from Transformers (BERT), a state-of-the-art deep learning approach. We apply BERT to analyze the texts of mandatory Management Discussion and Analysis (MD&A) sections of annual reports for a sample of 22,669 firm-year observations from 3,181 unique …

Non-Monotonic Generation Of Knowledge Paths For Context Understanding, Pei-Chi Lo, Ee-Peng Lim

Research Collection School Of Computing and Information Systems

Knowledge graphs can be used to enhance text search and access by augmenting textual content with relevant background knowledge. While many large knowledge graphs are available, using them to make semantic connections between entities mentioned in the textual content remains to be a difficult task. In this work, we therefore introduce contextual path generation (CPG) which refers to the task of generating knowledge paths, contextual path, to explain the semantic connections between entities mentioned in textual documents with given knowledge graph. To perform CPG task well, one has to address its three challenges, namely path relevance, incomplete knowledge graph, and …

Hypergraphs With Attention On Reviews For Explainable Recommendation, Theis E. Jendal, Trung Hoang Le, Hady Wirawan Lauw, Matteo Lissandrini, Peter Dolog, Katja Hose

Research Collection School Of Computing and Information Systems

Given a recommender system based on reviews, the challenges are how to effectively represent the review data and how to explain the produced recommendations. We propose a novel review-specific Hypergraph (HG) model, and further introduce a model-agnostic explainability module. The HG model captures high-order connections between users, items, aspects, and opinions while maintaining information about the review. The explainability module can use the HG model to explain a prediction generated by any model. We propose a path-restricted review-selection method biased by the user preference for item reviews and propose a novel explanation method based on a review graph. Experiments on …

Revisiting The Markov Property For Machine Translation, Cunxiao Du, Hao Zhou, Zhaopeng Tu, Jing Jiang

Research Collection School Of Computing and Information Systems

In this paper, we re-examine the Markov property in the context of neural machine translation. We design a Markov Autoregressive Transformer (MAT) and undertake a comprehensive assessment of its performance across four WMT benchmarks. Our findings indicate that MAT with an order larger than 4 can generate translations with quality on par with that of conventional autoregressive transformers. In addition, counter-intuitively, we also find that the advantages of utilizing a higher-order MAT do not specifically contribute to the translation of longer sentences.

Meta-Interpretive Learning With Reuse, Rong Wang, Jun Sun, Cong Tian, Zhenhua Duan

Research Collection School Of Computing and Information Systems

Inductive Logic Programming (ILP) is a research field at the intersection between machine learning and logic programming, focusing on developing a formal framework for inductively learning relational descriptions in the form of logic programs from examples and background knowledge. As an emerging method of ILP, Meta-Interpretive Learning (MIL) leverages the specialization of a set of higher-order metarules to learn logic programs. In MIL, the input includes a set of examples, background knowledge, and a set of metarules, while the output is a logic program. MIL executes a depth-first traversal search, where its program search space expands polynomially with the number …

Ur2m: Uncertainty And Resource-Aware Event Detection On Microcontrollers, Hong Jia, Young D. Kwon, Dong Ma, Nhat Pham, Lorena Qendro, Tam Vu, Cecilia Mascolo

Research Collection School Of Computing and Information Systems

Traditional machine learning techniques are prone to generating inaccurate predictions when confronted with shifts in the distribution of data between the training and testing phases. This vulnerability can lead to severe consequences, especially in applications such as mobile healthcare. Uncertainty estimation has the potential to mitigate this issue by assessing the reliability of a model's output. However, existing uncertainty estimation techniques often require substantial computational resources and memory, making them impractical for implementation on microcontrollers (MCUs). This limitation hinders the feasibility of many important on-device wearable event detection (WED) applications, such as heart attack detection. In this paper, we present …

Identification Of Faults In Highways Using Approximation Methods And Algorithms, Khudayberdiyev Khakkulmirzayevich Mirzaakbar, Anvar Asatilloyevich Ravshanov

Chemical Technology, Control and Management

Many fast Fourier transforms are used to identify defective parts of uneven surfaces on roads and send information to relevant organizations on the road, using the " RAVON YO‘LLAR" application installed on a mobile device during car movement. We determine the uneven parts of the road. Smooth and well-maintained roads reduce the risk of vehicle collisions, skidding and other road-related incidents. Timely measures contribute to overall safety, comfort and economic efficiency.

Efficiency Optimization Method For Data Sampling In Power Grid Topology Scheduling Simulation, Yingying Zhao, Pusen Dong, Tianchen Zhu, Fan Li, Yun Su, Zhenying Tai, Qingyun Sun, Hang Fan

Journal of System Simulation

Abstract: To address the large simulation computational workload and low simulation speed caused by the scale and complexity of the new power system, a simulation acceleration method for topology scheduling based on the distributed and quantization mechanisms is proposed. The parallelization of topology scheduling models is used to increase the scale of data simulation sampling in unit time. The introduced quantization operators accelerate the computation speed of the topology scheduling model operators, reduces the time cost of the every single simulation. Case studies confirm the effectiveness of the topology simulation acceleration, in which the available transfer capacity of the simulated …

Runoff Intelligent Prediction Method Based On Broad-Deep Fusion Time-Frequency Analysis, Ying Han, Lehao Wang, Shumei Wang, Xiang Zhang, Xingxing Luo

Journal of System Simulation

Abstract: Broad learning system(BLS) is introduced to tackle the existed disadvantage that LSTM-based runoff prediction model is easy to fall into local optimization. To reduce the influence of noise on the prediction results, the variational mode decomposition (VMD) is adopted to transform the onedimensional time-domain runoff signal to the two-dimensional time-frequency plane. The runoff prediction model based on VMD-LSTM-BLS is proposed. The simulation results demonstrate that the prediction accuracy of the new model is more significantly improved compared with the baseline model and the existing LSTM-based runoff prediction model.

Simulation Platform Of Agv System Scheduling Algorithms In Uncertain Environment, Zhihao Shi, Haihui Shen

Journal of System Simulation

Abstract: As a complete automated guided vehicle(AGV) system scheduling algorithm, it must include conflict-handling strategy, together with common dispatching strategy and routing algorithm. However, due to the uncertainty, characteristics of such scheduling algorithm are difficult to be analyzed theoretically, and the relevant study is lacking. An AGV system simulation platform based on the discreteevent simulation technique is designed and developed, which can flexibly set the scheduling problem, choose the dispatching strategy, routing algorithm, and conflict-handling strategy for the scheduling algorithm, to run the simulation. The platform has a visual interface, from which the running status of AGVs and the performance …

Research On Motion Planning Of Hexapod Robot Based On Drl And Free Gait, Xinpeng Wang, Huiqiao Fu, Guizhou Deng, Kaiqiang Tang, Chunlin Chen, Canghao Liu

Journal of System Simulation

Abstract: To improve the passability and the motion performance of the hexapod robot in the unstructured environment, a multi-contact motion planning algorithm based on DRL and free gait planner is proposed. Firstly, the free gait planner obtains the reachable footholds under the target state and outputs the optimal gait sequence. The center of mass motion policy of the hexapod robot in the randomly generated plum blossom pile environment is obtained by using deep reinforcement learning training. To ensure the reachability between adjacent states of the robot in motion, the state transition feasibility model is used to judge the state transition …

Robust Predictive Control Of Nonplanar Fully-Actuated Uavs, Yun Ma, Yuan Wang, Meng Li, Peng Wang, Yanling Tang

Journal of System Simulation

Abstract: Targeting the problem that nonplanar fully-actuated unmanned aerial vehicles (UAVs) are susceptible to external winds and unmodeled dynamics, the predictive control system with good robustness is designed. A nonlinear motion model with six degrees of freedom is established through the Newton-Euler approach. A linear extended state observer is designed to estimate the state variables by transforming the system affected by matched and unmatched disturbances into an equivalent system only affected by the matched disturbances. A predictive controller is designed for the equivalent system to reduce the output oscillation and input surging and a disturbance compensator is also designed to …

Reconnaissance Mission Planning Method For Air-Ground Heterogeneous Unmanned Systems, Guohui Zhang, Ya'nan Zhang, Ang Gao, Aoyu Xu

Journal of System Simulation

Abstract: Compared with the air-based homogeneous unmanned system, the motion capabilities, resource payloads, and combat scenes in the air-ground heterogeneous unmanned system increase the number of constraint conditions and significantly increase the computational complexity of the solution model. The modeling of collaborative combat missions and the efficient solution of large-scale problems are the key issues. With the time, path cost, and reconnaissance benefit as the objective functions, considering the constraints such as the endurance of unmanned platforms, a multi-objective programming model for the reconnaissance missions of an air-ground heterogeneous unmanned system is constructed. Aiming at the urban combat environments with …

A Simplification Method Of Large-Scale Unit Commitment Model Based On Boundary Method, Yanping Xu, Mingxin Zhao, Xiaohui Qin, Keyou He, Xiaohan Wu, Pei Zhang

Journal of System Simulation

Abstract: As the scale of power grid expands, in the market environment, the variables and constraints in the security-constrained unit commitment (SCUC) model considering power grid security constraints increase significantly and the solvability of the model reduces. When the model scale is too large, even the existing commercial solvers cannot solve it. Aiming at the model rapid solving, from the perspective of reducing the number of model constraints, a linear constraint simplification method based on the boundary method is proposed. The proposed method can effectively reduce the model's size by eliminating the redundant linear constraints. The IEEE-39, WECC 179 and …

Optimal Scheduling Strategy Of Virtual Power Plant With Carbon Emission And Carbon Penalty Considering Uncertainty Of Wind Power And Photovoltaic Power, Jijun Shui, Daogang Peng, Yankan Song, Qiang Zhou

Journal of System Simulation

Abstract: To better meet the development needs of China's new power system, an optimal scheduling strategy of virtual power plant(VPP) with carbon emission and carbon penalty considering the uncertainty of wind power and photovoltaic power is proposed. The mathematical description of photovoltaic(PV), wind turbine(WT), combined heat and power(CHP) unit and energy storage system (ESS) is carried out, and a wind-solar output model considering the uncertainty is established. The scenario generation and reduction method is used to generate the typical scenario. To maximize the overall operation benefit of VPP, considering carbon emission cost and carbon penalty, an optimal scheduling model of …