Physical Sciences and Mathematics Commons^™

Stochastic Gradient Hamiltonian Monte Carlo With Variance Reduction For Bayesian Inference, Zhize Li, Tianyi Zhang, Shuyu Cheng, Jun Zhu, Jian Li

Research Collection School Of Computing and Information Systems

Gradient-based Monte Carlo sampling algorithms, like Langevin dynamics and Hamiltonian Monte Carlo, are important methods for Bayesian inference. In large-scale settings, full-gradients are not affordable and thus stochastic gradients evaluated on mini-batches are used as a replacement. In order to reduce the high variance of noisy stochastic gradients, Dubey et al. (in: Advances in neural information processing systems, pp 1154–1162, 2016) applied the standard variance reduction technique on stochastic gradient Langevin dynamics and obtained both theoretical and experimental improvements. In this paper, we apply the variance reduction tricks on Hamiltonian Monte Carlo and achieve better theoretical convergence results compared with …

Compositional Coding For Collaborative Filtering, Chenghao Liu, Tao Lu, Xin Wang, Zhiyong Cheng, Jianling Sun, Steven C. H. Hoi

Research Collection School Of Computing and Information Systems

Efficiency is crucial to the online recommender systems, especially for the ones which needs to deal with tens of millions of users and items. Because representing users and items as binary vectors for Collaborative Filtering (CF) can achieve fast user-item affinity computation in the Hamming space, in recent years, we have witnessed an emerging research effort in exploiting binary hashing techniques for CF methods. However, CF with binary codes naturally suffers from low accuracy due to limited representation capability in each bit, which impedes it from modeling complex structure of the data. In this work, we attempt to improve the …

Outcasts – In Search Of Identity, Syed Hasan Haider

MSJ Capstone Projects

The idea for this documentary came from a story published in the express tribune which talked about the people who are unable to vote in 2018 elections due to having Computerized National Identity Cards (CNICs) in the Ibrahim Hyderi locality in Karachi.

Not having a CNIC in Pakistan means that you are not able to participate in civic life and also not subscribe to basic facilitates like housing, water, gas and employment.

This documentary film looks at different cases and through the experience of some journalists what it is like to live as an undocumented citizen. The film also explores …

A Hidden Markov Model For Matching Spatial Networks, Benoit Costes, Julien Perret

Journal of Spatial Information Science

Datasets of the same geographic space at different scales and temporalities are increasingly abundant, paving the way for new scientific research. These datasets require data integration, which implies linking homologous entities in a process called data matching that remains a challenging task, despite a quite substantial literature, because of data imperfections and heterogeneities. In this paper, we present an approach for matching spatial networks based on a hidden Markov model (HMM) that takes full benefit of the underlying topology of networks. The approach is assessed using four heterogeneous datasets (streets, roads, railway, and hydrographic networks), showing that the HMM algorithm …

Evaluating Existing Manually Constructed Natural Landscape Classification With A Machine Learning-Based Approach, Rok Ciglic, Erik Strumbelj, Rok Cesnovar, Mauro Hrvatin, Drago Perko

Journal of Spatial Information Science

Some landscape classifications officially determine financial obligations; thus, they must be objective and precise. We presume it is possible to quantitatively evaluate existing manually constructed classifications and correct them if necessary. One option for achieving this goal is a machine learning method. With (re)modeling of the landscape classification and an explanation of its structure, we can add quantitative proof to its original (qualitative) description. The main objectives of the paper are to evaluate the consistency of the existing manually constructed natural landscape classification with a machine learning-based approach and to test the newly developed general black-box explanation method in order …

Discovery Of Topological Constraints On Spatial Object Classes Using A Refined Topological Model, Ivan Majic, Elham Naghizade, Stephan Winter, Martin Tomko

Journal of Spatial Information Science

In a typical data collection process, a surveyed spatial object is annotated upon creation, and is classified based on its attributes. This annotation can also be guided by textual definitions of objects. However, interpretations of such definitions may differ among people, and thus result in subjective and inconsistent classification of objects. This problem becomes even more pronounced if the cultural and linguistic differences are considered. As a solution, this paper investigates the role of topology as the defining characteristic of a class of spatial objects. We propose a data mining approach based on frequent itemset mining to learn patterns in …

Data Mining And Machine Learning To Improve Northern Florida’S Foster Care System, Daniel Oldham, Nathan Foster, Mihhail Berezovski

Beyond: Undergraduate Research Journal

The purpose of this research project is to use statistical analysis, data mining, and machine learning techniques to determine identifiable factors in child welfare service records that could lead to a child entering the foster care system multiple times. This would allow us the capability of accurately predicting a case’s outcome based on these factors. We were provided with eight years of data in the form of multiple spreadsheets from Partnership for Strong Families (PSF), a child welfare services organization based in Gainesville, Florida, who is contracted by the Florida Department for Children and Families (DCF). This data contained a …

Healthcare It In Skilled Nursing And Post-Acute Care Facilities: Reducing Hospital Admissions And Re-Admissions, Improving Reimbursement And Improving Clinical Operations, Scott L. Hopes

Scott Hopes

Health information technology (HIT), which includes electronic health record (EHR) systems and clinical data analytics, has become a major component of all health care delivery and care management. The adoption of HIT by physicians, hospitals, post-acute care organizations, pharmacies and other health care providers has been accepted as a necessary (and recently, a government required) step toward improved quality, care coordination and reduced costs: “Better coordination of care provides a path to improving communication, improving quality of care, and reducing unnecessary emergency room use and hospital readmissions. LTPAC providers play a critical role in achieving these goals” (HealthIT.gov, 2013).

Though …

Unsupervised Deep Structured Semantic Models For Commonsense Reasoning, Shuohang Wang, Sheng Zhang, Yelong Shen, Xiaodong Liu, Jingjing Liu, Jianfeng Gao, Jing Jiang

Research Collection School Of Computing and Information Systems

Commonsense reasoning is fundamental to natural language understanding. While traditional methods rely heavily on human-crafted features and knowledge bases, we explore learning commonsense knowledge from a large amount of raw text via unsupervised learning. We propose two neural network models based on the Deep Structured Semantic Models (DSSM) framework to tackle two classic commonsense reasoning tasks, Winograd Schema challenges (WSC) and Pronoun Disambiguation (PDP). Evaluation shows that the proposed models effectively capture contextual information in the sentence and co-reference information between pronouns and nouns, and achieve significant improvement over previous state-of-the-art approaches.

Spatio-Temporal Multimedia Big Data Analytics Using Deep Neural Networks, Samira Pouyanfar

FIU Electronic Theses and Dissertations

With the proliferation of online services and mobile technologies, the world has stepped into a multimedia big data era, where new opportunities and challenges appear with the high diversity multimedia data together with the huge amount of social data. Nowadays, multimedia data consisting of audio, text, image, and video has grown tremendously. With such an increase in the amount of multimedia data, the main question raised is how one can analyze this high volume and variety of data in an efficient and effective way. A vast amount of research work has been done in the multimedia area, targeting different aspects …

Multimodal Data Analytics And Fusion For Data Science, Haiman Tian

FIU Electronic Theses and Dissertations

Advances in technologies have rapidly accumulated a zettabyte of “new” data every two years. The huge amount of data have a powerful impact on various areas in science and engineering and generates enormous research opportunities, which calls for the design and development of advanced approaches in data analytics. Given such demands, data science has become an emerging hot topic in both industry and academia, ranging from basic business solutions, technological innovations, and multidisciplinary research to political decisions, urban planning, and policymaking. Within the scope of this dissertation, a multimodal data analytics and fusion framework is proposed for data-driven knowledge discovery …

Reach - A Community Service Application, Samuel Noel Magana

Computer Engineering

Communities are familiar threads that unite people through several shared attributes and interests. These commonalities are the core elements that link and bond us together. Many of us are part of multiple communities, moving in and out of them depending on our needs. These common threads allow us to support and advocate for each other when facing a common threat or difficult situation. Healthy and vibrant communities are fundamental to the operation of our society. These interactions within our communities define the way we as individuals interact with each other, and society at large. Being part of a community helps …

Keylime, Eli William Partker

Computer Engineering

Josh, Matt and I knew we wanted to develop a mobile app for our senior project because that is what we found ourselves to be most passionate about during our time here at Cal Poly. We started to think of problems we wanted to solve using an application and we came up with a couple ideas but chose to expand on one. Students come to Cal Poly every year new to the area and the food options San Luis Obispo provides. Many of the restaurants in SLO offer a variety of deals to the community and most of them to …

Radish: A Cross Platform Meal Prepping App For Beginner Weightlifters, Spoorthy S. Vemula, Tanay Gottigundala, Cory Baxes

Computer Science and Software Engineering

With the increasing ease of access and decreasing price of most food, obesity rates in the developing world have risen dramatically in recent years. As of March 23rd, 2019, obesity rates had reached 39.6%, a 6% increase in just 8 years. Research has shown that people with obesity have a significantly increased risk of heart disease, stroke, type 2 diabetes, and certain cancers, among other life-threatening diseases. In addition, 42% of people who begin weightlifting quit because it’s too difficult to follow a diet or workout regimen.

We created Radish in an attempt to tackle these problems. Radish makes it …

Salient Object Detection With Pyramid Attention And Salient Edges, Wenguan Wang, Shuyang Zhao, Jianbing Shen, Steven C. H. Hoi, Ali Borji

Research Collection Yong Pung How School Of Law

This paper presents a new method for detecting salient objects in images using convolutional neural networks (CNNs). The proposed network, named PAGE-Net, offers two key contributions. The first is the exploitation of an essential pyramid attention structure for salient object detection. This enables the network to concentrate more on salient regions while considering multi-scale saliency information. Such a stacked attention design provides a powerful tool to efficiently improve the representation ability of the corresponding network layer with an enlarged receptive field. The second contribution lies in the emphasis on the importance of salient edges. Salient edge information offers a strong …

Metagraph-Based Learning On Heterogeneous Graphs, Yuan Fang, Wenqing Lin, Vincent W. Zheng, Min Wu, Jiaqi Shi, Kevin Chang, Xiao-Li Li

Research Collection School Of Computing and Information Systems

Data in the form of graphs are prevalent, ranging from biological and social networks to citation graphs and the Web. Inparticular, most real-world graphs are heterogeneous, containing objects of multiple types, which present new opportunities for manyproblems on graphs. Consider a typical proximity search problem on graphs, which boils down to measuring the proximity between twogiven nodes. Most earlier studies on homogeneous or bipartite graphs only measure a generic form of proximity, without accounting fordifferent “semantic classes”—for instance, on a social network two users can be close for different reasons, such as being classmates orfamily members, which represent two distinct …

Dynamic Fusion With Intra-And Inter-Modality Attention Flow For Visual Question Answering, Peng Gao, Zhengkai Jiang, Haoxuan You, Pan Lu, Steven C. H. Hoi, Xiaogang Wang, Hongsheng Li

Research Collection School Of Computing and Information Systems

Learning effective fusion of multi-modality features is at the heart of visual question answering. We propose a novel method of dynamically fusing multi-modal features with intra- and inter-modality information flow, which alternatively pass dynamic information between and across the visual and language modalities. It can robustly capture the high-level interactions between language and vision domains, thus significantly improves the performance of visual question answering. We also show that the proposed dynamic intra-modality attention flow conditioned on the other modality can dynamically modulate the intramodality attention of the target modality, which is vital for multimodality feature fusion. Experimental evaluations on the …

Learning Unsupervised Video Object Segmentation Through Visual Attention, Wenguan Wang, Hongmei Song, Shuyang Zhao, Jianbing Shen, Sanyuan Zhao, Steven C. H. Hoi, Haibin Ling

Research Collection Yong Pung How School Of Law

This paper conducts a systematic study on the role of visual attention in Unsupervised Video Object Segmentation (UVOS) tasks. By elaborately annotating three popular video segmentation datasets (DAVIS, Youtube-Objects and SegTrack V2) with dynamic eye-tracking data in the UVOS setting, for the first time, we quantitatively verified the high consistency of visual attention behavior among human observers, and found strong correlation between human attention and explicit primary object judgements during dynamic, task-driven viewing. Such novel observations provide an in-depth insight into the underlying rationale behind UVOS. Inspired by these findings, we decouple UVOS into two sub-tasks: UVOS-driven Dynamic Visual Attention …

Matching Passengers And Drivers With Multiple Objectives In Ride Sharing Markets, Guodong Lyu, Chung Piaw Teo, Wangchi Cheung, Hai Wang

Research Collection School Of Computing and Information Systems

In many cities in the world, ride sharing companies, such as Uber, Didi, Grab and Lyft, have been able to leverage on Internet-based platforms to conduct online decision making to connect passengers and drivers. These online platforms facilitate the integration of passengers and drivers’ mobility data on smart phones in real-time, which enables a convenient matching between demand and supply in real time. These clear operational advantages have motivated many similar shared service business models in the public transportation arena, and have been a disruptive force to the traditional taxi industry.

Distributed Similarity Queries In Metric Spaces, Keyu Yang, Xin Ding, Yuanliang Zhang, Lu Chen, Baihua Zheng, Yunjun Gao

Research Collection School Of Computing and Information Systems

Similarity queries, including range queries and k nearest neighbor (kNN) queries, in metric spaces have applications in many areas such as multimedia retrieval, computational biology and location-based services. With the growing volumes of data, a distributed method is required. In this paper, we propose an Asynchronous Metric Distributed System (AMDS), to support efficient metric similarity queries in the distributed environment. AMDS uniformly partitions the data with the pivot-mapping technique to ensure the load balancing, and employs publish/subscribe communication model to asynchronous process large scale of queries. The employment of asynchronous processing model also improves robustness and efficiency of AMDS. In …

Sliced Wasserstein Generative Models, Jiqing Wu, Zhiwu Huang, Dinesh Acharya, Wen Li, Janine Thoma, Danda Pani Paudel, Luc Van Gool

Research Collection School Of Computing and Information Systems

In generative modeling, the Wasserstein distance (WD) has emerged as a useful metric to measure the discrepancy between generated and real data distributions. Unfortunately, it is challenging to approximate the WD of high-dimensional distributions. In contrast, the sliced Wasserstein distance (SWD) factorizes high-dimensional distributions into their multiple one-dimensional marginal distributions and is thus easier to approximate. In this paper, we introduce novel approximations of the primal and dual SWD. Instead of using a large number of random projections, as it is done by conventional SWD approximation methods, we propose to approximate SWDs with a small number of parameterized orthogonal projections …

Dietlens-Eout: Large Scale Restaurant Food Photo Recognition, Zhipeng Wei, Jingjing Chen, Zhaoyan Ming, Chong-Wah Ngo, Tat-Seng Chua, Fengfeng Zhou

Research Collection School Of Computing and Information Systems

Restaurant dishes represent a significant portion of food that people consume in their daily life. While people are becoming healthconscious in their food intake, convenient restaurant food tracking becomes an essential task in wellness and fitness applications. Given the huge number of dishes (food categories) involved, it becomes extremely challenging for traditional food photo classification to be feasible in both algorithm design and training data availability. In this work, we present a demo that runs on restaurant dish images in a city of millions of residents and tens of thousand restaurants. We propose a rank-loss based convolutional neural network to …

Mixed Dish Recognition Through Multi-Label Learning, Yunan Wang, Jing-Jing Chen, Chong-Wah Ngo, Tat-Seng Chua, Wanli Zuo, Zhaoyan Ming

Research Collection School Of Computing and Information Systems

Mix dish recognition, whose goal is to identify each of the dish type presented on one plate, is generally regarded as a difficult problem. The major challenge of this problem is that different dishes presented in one plate may overlap with each other and there may be no clear boundaries among them. Therefore, labeling the bounding box of each dish type is difficult and not necessarily leading to good results. This paper studies the problem from the perspective of multi-label learning. Specially, we propose to perform dish recognition on region level with multiple granularities. For experimental purpose, we collect two …

Self-Supervised Spatio-Temporal Representation Learning For Videos By Predicting Motion And Appearance Statistics, Jiangliu Wang, Jianbo Jiao, Linchao Bao, Shengfeng He, Yunhui Liu, Wei Liu

Research Collection School Of Computing and Information Systems

We address the problem of video representation learning without human-annotated labels. While previous efforts address the problem by designing novel self-supervised tasks using video data, the learned features are merely on a frame-by-frame basis, which are not applicable to many video analytic tasks where spatio-temporal features are prevailing. In this paper we propose a novel self-supervised approach to learn spatio-temporal features for video representation. Inspired by the success of two-stream approaches in video classification, we propose to learn visual features by regressing both motion and appearance statistics along spatial and temporal dimensions, given only the input video data. Specifically, we …

View, Like, Comment, Post: Analyzing User Engagement By Topic At 4 Levels Across 5 Social Media Platforms For 53 News Organizations, Kholoud K. Aldous, Jisun An, Bernard J. Jansen

Research Collection School Of Computing and Information Systems

We evaluate the effects of the topics of social media posts on audiences across five social media platforms (i.e., Facebook, Instagram, Twitter, YouTube, and Reddit) at four levels of user engagement. We collected 3,163,373 social posts from 53 news organizations across five platforms during an 8month period. We analyzed the differences in news organization platform strategies by focusing on topic variations by organization and the corresponding effect on user engagement at four levels. Findings show that topic distribution varies by platform, although there are some topics that are popular across most platforms. User engagement levels vary both by topics and …

Time-Based Payout Ratio For Coordinating Supply And Demand On An On-Demand Service Platform, Jiaru Bai, Kut C. So, Christopher S. Tang, Xiqun Chen, Hai Wang

Research Collection School Of Computing and Information Systems

Many on-demand service platforms use a fixed payout ratio (i.e., the percentage of the platform’s revenue that is paid to the providers) regardless of the customer demand and the number of participating providers that tend to vary over time. In this chapter, we examine the implications of time-based payout ratios. To do so, we first present a queueing model with endogenous supply (number of participating providers) and endogenous demand (customer request rate) to model this on-demand service platform. In our model, earnings-sensitive independent providers have heterogeneous reservation price (for work participation) to serve wait-time and price-sensitive customers with heterogeneous valuation …

Learning Cross-Modal Embeddings With Adversarial Networks For Cooking Recipes And Food Images, Hao Wang, Doyen Sahoo, Chenghao Liu, Ee-Peng Lim, Steven C. H. Hoi

Research Collection School Of Computing and Information Systems

Food computing is playing an increasingly important role in human daily life, and has found tremendous applications in guiding human behavior towards smart food consumption and healthy lifestyle. An important task under the food-computing umbrella is retrieval, which is particularly helpful for health related applications, where we are interested in retrieving important information about food (e.g., ingredients, nutrition, etc.). In this paper, we investigate an open research task of cross-modal retrieval between cooking recipes and food images, and propose a novel framework Adversarial Cross-Modal Embedding (ACME) to resolve the cross-modal retrieval task in food domains. Specifically, the goal is to …

Context-Aware Spatio-Recurrent Curvilinear Structure Segmentation, Feigege Wang, Yue Gu, Wenxi Liu, Shengfeng He, Shengfeng He, Jia Pan

Research Collection School Of Computing and Information Systems

Curvilinear structures are frequently observed in various images in different forms, such as blood vessels or neuronal boundaries in biomedical images. In this paper, we propose a novel curvilinear structure segmentation approach using context-aware spatio-recurrent networks. Instead of directly segmenting the whole image or densely segmenting fixed-sized local patches, our method recurrently samples patches with varied scales from the target image with learned policy and processes them locally, which is similar to the behavior of changing retinal fixations in the human visual system and it is beneficial for capturing the multi-scale or hierarchical modality of the complex curvilinear structures. In …

Stabilized Svrg: Simple Variance Reduction For Nonconvex Optimization, Rong Ge, Zhize Li, Weiyao Wang, Xiang Wang

Research Collection School Of Computing and Information Systems

Variance reduction techniques like SVRG provide simple and fast algorithms for optimizing a convex finite-sum objective. For nonconvex objectives, these techniques can also find a first-order stationary point (with small gradient). However, in nonconvex optimization it is often crucial to find a second-order stationary point (with small gradient and almost PSD hessian). In this paper, we show that Stabilized SVRG (a simple variant of SVRG) can find an $\epsilon$-second-order stationary point using only $\tilde{O}(n^{2/3}/\epsilon^2 + n/\epsilon^{1.5})$ stochastic gradients. To our best knowledge, this is the first second-order guarantee for a simple variant of SVRG. The running time almost matches the …

A Study Of Machine Learning And Deep Learning Models For Solving Medical Imaging Problems, Fadi G. Farhat

Theses

Application of machine learning and deep learning methods on medical imaging aims to create systems that can help in the diagnosis of disease and the automation of analyzing medical images in order to facilitate treatment planning. Deep learning methods do well in image recognition, but medical images present unique challenges. The lack of large amounts of data, the image size, and the high class-imbalance in most datasets, makes training a machine learning model to recognize a particular pattern that is typically present only in case images a formidable task.

Experiments are conducted to classify breast cancer images as healthy or …