Digital Commons Network^™

Acamda: Improving Data Efficiency In Reinforcement Learning Through Guided Counterfactual Data Augmentation, Yuewen Sun, Erli Wang, Biwei Huang, Chaochao Lu, Lu Feng, Changyin Sun, Kun Zhang

Machine Learning Faculty Publications

Data augmentation plays a crucial role in improving the data efficiency of reinforcement learning (RL). However, the generation of high-quality augmented data remains a significant challenge. To overcome this, we introduce ACAMDA (Adversarial Causal Modeling for Data Augmentation), a novel framework that integrates two causality-based tasks: causal structure recovery and counterfactual estimation. The unique aspect of ACAMDA lies in its ability to recover temporal causal relationships from limited non-expert datasets. The identification of the sequential cause-and-effect allows the creation of realistic yet unobserved scenarios. We utilize this characteristic to generate guided counterfactual datasets, which, in turn, substantially reduces the need …

Dynamic Spiking Graph Neural Networks, Nan Yin, Mengzhu Wang, Zhenghan Chen, Giulia De Masi, Huan Xiong, Bin Gu

Machine Learning Faculty Publications

The integration of Spiking Neural Networks (SNNs) and Graph Neural Networks (GNNs) is gradually attracting attention due to the low power consumption and high efficiency in processing the non-Euclidean data represented by graphs. However, as a common problem, dynamic graph representation learning faces challenges such as high complexity and large memory overheads. Current work often uses SNNs instead of Recurrent Neural Networks (RNNs) by using binary features instead of continuous ones for efficient training, which would overlooks graph structure information and leads to the loss of details during propagation. Additionally, optimizing dynamic spiking models typically requires propagation of information across …

Enhancing Training Of Spiking Neural Network With Stochastic Latency, Srinivas Anumasa, Bhaskar Mukhoty, Velibor Bojković, Giulia De Masi, Huan Xiong, Bin Gu

Machine Learning Faculty Publications

Spiking neural networks (SNNs) have garnered significant attention for their low power consumption when deployed on neuromorphic hardware that operates in orders of magnitude lower power than general-purpose hardware. Direct training methods for SNNs come with an inherent latency for which the SNNs are optimized, and in general, the higher the latency, the better the predictive powers of the models, but at the same time, the higher the energy consumption during training and inference. Furthermore, an SNN model optimized for one particular latency does not necessarily perform well in lower latencies, which becomes relevant in scenarios where it is necessary …

Exploring Channel-Aware Typical Features For Out-Of-Distribution Detection, Rundong He, Yue Yuan, Zhongyi Han, Fan Wang, Wan Su, Yilong Yin, Tongliang Liu, Yongshun Gong

Machine Learning Faculty Publications

Detecting out-of-distribution (OOD) data is essential to ensure the reliability of machine learning models when deployed in real-world scenarios. Different from most previous test-time OOD detection methods that focus on designing OOD scores, we delve into the challenges in OOD detection from the perspective of typicality and regard the feature’s high-probability region as the feature’s typical set. However, the existing typical-feature-based OOD detection method implies an assumption: the proportion of typical feature sets for each channel is fixed. According to our experimental analysis, each channel contributes differently to OOD detection. Adopting a fixed proportion for all channels results in several …

Iterative Regularization With K-Support Norm: An Important Complement To Sparse Recovery, William De Vazelhes, Bhaskar Mukhoty, Xiao Tong Yuan, Bin Gu

Machine Learning Faculty Publications

Sparse recovery is ubiquitous in machine learning and signal processing. Due to the NP-hard nature of sparse recovery, existing methods are known to suffer either from restrictive (or even unknown) applicability conditions, or high computational cost. Recently, iterative regularization methods have emerged as a promising fast approach because they can achieve sparse recovery in one pass through early stopping, rather than the tedious grid-search used in the traditional methods. However, most of those iterative methods are based on the ℓ1 norm which requires restrictive applicability conditions and could fail in many cases. Therefore, achieving sparse recovery with iterative regularization methods …

Limited Memory Online Gradient Descent For Kernelized Pairwise Learning With Dynamic Averaging, Hilal Alquabeh, William De Vazelhes, Bin Gu

Machine Learning Faculty Publications

Pairwise learning, an important domain within machine learning, addresses loss functions defined on pairs of training examples, including those in metric learning and AUC maximization. Acknowledging the quadratic growth in computation complexity accompanying pairwise loss as the sample size grows, researchers have turned to online gradient descent (OGD) methods for enhanced scalability. Recently, an OGD algorithm emerged, employing gradient computation involving prior and most recent examples, a step that effectively reduces algorithmic complexity to O(T), with T being the number of received examples. This approach, however, confines itself to linear models while assuming the independence of example arrivals. We introduce …

Robustly Train Normalizing Flows Via Kl Divergence Regularization, Kun Song, Ruben Solozabal, Hao Li, Martin Takáč, Lu Ren, Fakhri Karray

Machine Learning Faculty Publications

In this paper, we find that the training of Normalizing Flows (NFs) are easily affected by the outliers and a small number (or high dimensionality) of training samples. To solve this problem, we propose a Kullback–Leibler (KL) divergence regularization on the Jacobian matrix of NFs. We prove that such regularization is equivalent to adding a set of samples whose covariance matrix is the identity matrix to the training set. Thus, it reduces the negative influence of the outliers and the small sample number on the estimation of the covariance matrix, simultaneously. Therefore, our regularization makes the training of NFs robust. …

Knowledge-Aware Explainable Reciprocal Recommendation, Kai Huang Lai, Zhe Rui Yang, Pei Yuan Lai, Chang Dong Wang, Mohsen Guizani, Min Chen

Machine Learning Faculty Publications

Reciprocal recommender systems (RRS) have been widely used in online platforms such as online dating and recruitment. They can simultaneously fulfill the needs of both parties involved in the recommendation process. Due to the inherent nature of the task, interaction data is relatively sparse compared to other recommendation tasks. Existing works mainly address this issue through content-based recommendation methods. However, these methods often implicitly model textual information from a unified perspective, making it challenging to capture the distinct intentions held by each party, which further leads to limited performance and the lack of interpretability. In this paper, we propose a …

Prior And Prediction Inverse Kernel Transformer For Single Image Defocus Deblurring, Peng Tang, Zhiqiang Xu, Chunlai Zhou, Pengfei Wei, Peng Han, Xin Cao, Tobias Lasser

Machine Learning Faculty Publications

Defocus blur, due to spatially-varying sizes and shapes, is hard to remove. Existing methods either are unable to effectively handle irregular defocus blur or fail to generalize well on other datasets. In this work, we propose a divide-and-conquer approach to tackling this issue, which gives rise to a novel end-to-end deep learning method, called prior-and-prediction inverse kernel transformer (P2IKT), for single image defocus deblurring. Since most defocus blur can be approximated as Gaussian blur or its variants, we construct an inverse Gaussian kernel module in our method to enhance its generalization ability. At the same time, an inverse kernel prediction …

Skin-Former: Mobile-Friendly Transformer For Skin Lesion Diagnosis, Sheng Zhang, Muzammal Naseer, Guangyi Chen, Zhiqiang Shen, Salman Khan, Kun Zhang, Fahad Shahbaz Khan

Machine Learning Faculty Publications

Large-scale pre-trained Vision Language Models (VLMs) have proven effective for zero-shot classification. Despite the success, most traditional VLMs-based methods are restricted by the assumption of partial source supervision or ideal target vocabularies, which rarely satisfy the open-world scenario. In this paper, we aim at a more challenging setting, Realistic Zero-Shot Classification, which assumes no annotation but instead a broad vocabulary. To address the new problem, we propose the Self Structural Semantic Alignment (S3A) framework, which extracts the structural semantic information from unlabeled data while simultaneously self-learning. Our S3A framework adopts a unique Cluster-Vote-Prompt-Realign (CVPR) algorithm, which iteratively groups unlabeled data …

Book Reviews: Jane Goodall At 90 & The Emotional Lives Of Animals, 2024 Revision, Andrew Rowan

Tales of WellBeing

Jane Goodall and Marc Bekoff have made significant contributions to our understanding of animal emotions and feelings.

Investigating Channel Narrowing And Nonnative Vegetation Encroachment In Three Dryland Rivers, Casey Pennock, Benjamin Miller, Phaedra Budy

Browse all Datasets

Water development and the proliferation of invasive riparian vegetation has led to widespread habitat loss and simplification of rivers in the western United States, contributing to the imperilment of native fishes. Here, we quantify channel narrowing and vegetation encroachment, which are conspicuous indicators of riverine habitat degradation, along approximately 400 km of three dryland tributaries of the upper Colorado River. To accomplish this, we conducted a comparative analysis of aerial photographs from historical (1930s) and contemporary (2010s or 2020s) time periods and utilized Light Detection and Ranging (LiDAR) data and Object-Based Image Analysis (OBIA) to determine contemporary canopy cover of …

Using Nonnative Vegetation To Enhance In-Stream Habitat For Native Fishes, Casey Pennock, Benjamin Miller, Phaedra Budy

Browse all Datasets

Flow alteration and riparian vegetation encroachment are causing habitat simplification with severe consequences for native fishes. To assess the effectiveness of enhancing simplified habitat in a large dryland river, we experimentally added invasive wood at 19 paired treatment and reference (no wood added) subreaches (50 - 100m) within the main channel of the San Juan River. Using a before-after-control-impact design, we sampled fishes and macroinvertebrates, and quantified habitat complexity. After wood addition, total native fish densities were 2.2x higher in treatments compared to references, whereas total nonnative fish densities exhibited no response. Macroinvertebrate densities were 6.8x higher, and habitat complexity …

Risk Of Secondary Malignancies After Pelvic Radiation: A Population-Based Analysis, Connor Mcpartland, Andrew Salib, Joshua Banks, James R. Mark, Costas D. Lallas, Edouard J. Trabulsi, Leonard G. Gomella, Hannan Goldberg, Benjamin Leiby, Robert Den, Thenappan Chandrasekar

Department of Urology Faculty Papers

Background and objective

Radiation therapy has increasingly been used in the management of pelvic malignancies. However, the use of radiation continues to pose a risk of a secondary malignancy to its recipients. This study investigates the risk of secondary malignancy development following radiation for primary pelvic malignancies.

Methods

A retrospective cohort review of the Surveillance, Epidemiology, and End Results database from 1975 to 2016 was performed. Primary pelvic malignancies were subdivided based on the receipt of radiation, and secondary malignancies were stratified as pelvic or nonpelvic to investigate the local effect of radiation.

Key findings and limitations

A total of …

Technical Data Package For Sysmlv2 Vignettes, Allen W. Dukes

Student Publications

This Technical Data Package (TDP) thoroughly compiles instructions, concepts, and solutions for eight SysMLv2 vignettes. Each vignette illustrates methods to solve complex system modeling challenges. The initial concept for each vignette can drive the use and evaluation of a custom SysMLv2 modeling tool as shown in [1]. They can support a baseline set of tasks to compare usability with other SysMLv2 modeling software tools.

The document starts with three pairs of vignettes for modeling a hypothetical Unmanned Aerial Vehicle (UAV) operated with a Virtual Reality (VR) headset derived from AFIT WKSP 696. The final pair of vignettes focus on two …

Fluvial Acoustics Of Lucky Peak Dam, Jacob Anderson

Fluvial Acoustics Data

This dataset includes infrasound, audible sound, photos, and video recorded near the outflow of Lucky Peak Dam near Boise, ID.

Employment In High Needs Schools First Year Post-Graduation, Debbie Hahs-Vaughn, Mary Little, Christine Destefano, Oluwaseun Farotimi

Research Data and Datasets

The data were part of study from a five-year federally funded project conducted at a large public university in the southern United States in partnership with a local school district and select high needs schools within the district. As part of the university’s internship process, teacher candidates applying for their internship identify schools at which they prefer to complete their student teaching. Teacher candidates identified for this study were those who had completed all requirements for entering internship, who selected one or more of the partner high needs schools for internship placement, and who were ultimately assigned to one of …

"Success Is The Only Option", Sherene A. Carpenter Phd

National Youth Advocacy & Resilience Conference

"Success Is the Only Option". Reflective, Engaging, Imperative. Often times teachers place grades on report cards without analyzing or reflecting. Interesting conversations take place when teachers are presented with a chart displaying the number of As and Bs compared to the number Ds and Fs. What does a snapshot of your classroom, school, or district reveal about both student and teacher academic success? This presentation allows participants to identify resolutions to barriers, as well as receive tools that enhance student/teacher engagement - as Academic Success Is the Only Option.

Energy Optimization In Sustainable Smart Environments With Machine Learning And Advanced Communications, Lidia Bereketeab, Aymen Zekeria, Moayad Aloqaily, Mohsen Guizani, Merouane Debbah

Machine Learning Faculty Publications

Enhancing energy optimization is crucial for sustainable and smart environments such as smart cities, connected and urban buildings, and cognitive cities. Advanced communication systems and Internet of Things (IoT) sensor systems play a key role in enhancing energy efficiency by monitoring and controlling such ecosystems. In this article, we propose a reinforcement learning (RL) approach for optimizing the energy consumption of multipurpose buildings using the EnergyPlus simulation environment. Our RL algorithm uses the proximal policy optimization with clipping (PPO-Clip) for online training and also includes an offline pretraining model to improve the stability of the proposed algorithm. The observed states …

Design Of Protograph Ldpc-Coded Mimo-Vlc Systems With Generalized Spatial Modulation, Dai Lin, Fang Yi, Guan Yongliang, Mohsen Guizani

Machine Learning Faculty Publications

This paper investigates the bit-interleaved coded generalized spatial modulation (BICGSM) with iterative decoding (BICGSM-ID) for multiple-input multiple-output (MIMO) visible light communications (VLC). In the BICGSM-ID scheme, the information bits conveyed by the signal-domain (SiD) symbols and the spatial-domain (SpD) light emitting diode (LED)index patterns are coded by a protograph low-density parity-check (P-LDPC) code. Specifically, we propose a signal-domain symbol expanding and re-allocating (SSER) method for constructing a type of novel generalized spatial modulation (GSM) constellations, referred to as SSERGSM constellations, so as to boost the performance of the BICGSM-ID MIMO-VLC systems. Moreover, by applying a modified PEXIT (MPEXIT) algorithm, we …

University Of Missouri-St. Louis Open Educational Resources Open Dataset, Helena Marvin

UMSL Datasets

The UMSL OER Open Dataset is a collection of data pertaining to Affordable and Open Educational Resources (OER) courses offered at the University of Missouri-St. Louis (UMSL). The dataset includes information such as course names, IDs, instructors, meeting patterns, enrollment numbers, and more. This dataset is not exhaustive, but it does include a helpful data dictionary to help orient researchers, educators, and students to better utilize the dataset to explore potential student cost savings and patterns in affordable and OER course offerings at the University of Missouri-St. Louis.

Data visualizations of this data can be found at https://public.tableau.com/views/UMSLAOERSavingsOpenDataset/Dashboard1?:language=en-US&:sid=&:display_count=n&:origin=viz_share_link

Candy Costume Royal, Nala Scott

Game Design

A board game where player compete to be the best dressed on the board. Players race to make the best outfit and reach the end of the board first.

Why Is This Man's Urine Flaky?, Brittney Hulsey

PA Faculty Publications

No abstract provided.

Dataset For Observing The Effect Of Morphology On Elastic Properties Of New Snow Using A Non-Contacting Laser Ultrasound System, J. Chris Mccaslin, T. Dylan Mikesell, H.-P. Marshall, Zoe Courville

CryoGARS Snow Data

Quantifying the mechanical properties of snow is crucial for various applications. Snow density alone can provide a rough estimate of mechanical properties, while direct observation of snow microstructure is necessary to accurately determine mechanical properties. We utilize a novel non-contacting laser ultrasound system (LUS), providing acoustic waveform measurements, to observe mechanical properties at the microscale. We investigated the temporal relationship between P-wave velocity, density, snow crystal type, and specific surface area (SSA). We created homogeneous snow samples, each composed of a single crystal type, compacted to a 250~$kg/m^3$ density. We measured acoustic wave propagation through these snow samples to observe …

Green Oa Day 2024 Toolkit, Usu Libraries, Erica Finch

Library Faculty & Staff Publications

This toolkit contains graphic assets; social media images for Facebook, Instagram, and X; and a PowerPoint template developed by Utah State University to promote green open access to faculty, staff, and students.

Dna-Pkcs Rnaseq Data: Dna-Pkcs Wilde-Type Or Kinase-Dead Protein Regulate Basal And Etoposide-Induced Gene Expression Changes, Amanda Ashley

Chemistry & Biochemistry: Datasets

No abstract provided.

Role Of Gd3+ Ions In Alumino Borosilicate Glasses Studied Through Structural, Optical And Thermoluminescence Characteristics For Gamma Dosimetry Applications, M. Monisha, Rashmi Nambiar, K. R. Vighnesh, A. Vidya Saraswathi

Open Access archive

This paper reports the structural, optical and thermoluminescence properties of Gd³⁺ ions incorporated alumino-borosilicate glasses, prepared through the melt-quench technique. The amorphous nature of the glasses is confirmed through XRD study. Network vibrations of borate and silicate groups, vibrations of hydrogen bonding and hydroxyl groups are realized via FTIR study. The Gd³⁺ transition peaks are absent in the near ultraviolet, visible and NIR regions, and the indirect bandgap values were determined through UV–Visible–NIR study. Photoluminescence measurements showed a concentration quenching in the glasses beyond 1.5 mol % of Gd³⁺ ions. Based on the thermoluminescence (TL) data, the …

Consumer's Interaction With Chatgpt- A Utaut Perspective, K. Shilpa, Devadas Menon

Open Access archive

This research paper explores consumer's interaction with ChatGPT, the latest version of OpenAI's advanced language processing AI model. ChatGPT's ability to understand and respond to questions naturally has made it a rapidly growing application worldwide, surpassing even popular social media platforms. With fears of job displacement and concerns about unequal benefits from AI, understanding consumer engagement with ChatGPT becomes crucial. Using a qualitative framework, this study focuses on early adopters of ChatGPT in India, as their opinions and sentiments can shape the overall perception of this new technology. The Unified Theory of Acceptance and Use of Technology (UTAUT) model serves …

Unlocking The Promise Of Liquid Biopsies In Precision Oncology, Alejandra Pando-Caciano, Rakesh Trivedi, Jarne Pauwels, Joanna Nowakowska

Open Access archive

Liquid biopsies have emerged as a promising and minimally invasive alternative to traditional tissue biopsies for detecting and monitoring cancer. Liquid biopsies offer a comprehensive analysis of cancer genetics and tumor burden by examining circulating cells and cell-derived analytes using a variety of assays, including conventional PCR methods and cutting-edge tools like long-read sequencing and nanotechnology. However, there are still some limitations and challenges that need to be overcome for their implementation in clinical routine, including the need for further research on their sensitivity and specificity, cost-effectiveness, standardization, and regulatory approval. Despite these challenges, liquid biopsies have the potential to …

Global, Regional, And National Burden Of Neck Pain, 1990–2020, And Projections To 2050: A Systematic Analysis Of The Global Burden Of Disease Study 2021, Ai Min Wu, Marita Cross, James M. Elliott, Garland T. Culbreth

Open Access archive

Background: Neck pain is a highly prevalent condition that leads to considerable pain, disability, and economic cost. We present the most current estimates of neck pain prevalence and years lived with disability (YLDs) from the Global Burden of Diseases, Injuries, and Risk Factors Study (GBD) by age, sex, and location, with forecasted prevalence to 2050. Methods: Systematic reviews identified population-representative surveys used to estimate the prevalence of and YLDs from neck pain in 204 countries and territories, spanning from 1990 to 2020, with additional data from opportunistic review. Medical claims data from Taiwan (province of China) were also included. Input …