Total: 16
The number of images is growing rapidly in recent years because of development of Internet, especially the social networks like Facebook, and the popularization of portable image capture devices like smart phone. Annotating them with semantically meaningful words to describe them, i.e., classification, is a useful way to manage these images. However, the huge number of images and classes brings several challenges to classification, of which two are 1) how to measure the similarity efficiently between large-scale images, for example, measuring similarity between samples is the building block for SVM and kNN classifiers, and 2) how to train supervised classification models for newly emerging classes with only a few or even no labeled samples because new concepts appear every day in the Web, like Tesla's Model S. The research of my Ph. D. thesis focuses on the two problems in large-scale image classification mentioned above. Formally, these two problems are termed as large-scale similarity search which focuses on the large scale of samples/images and zero-shot/few-shots learning which focuses on the large scale of classes. Specifically, my research considers the following three aspects: 1) hashing based large-scale similarity search which adopts hashing to improve the efficiency; 2) cross-class transfer active learning which simultaneously transfers knowledge from the abundant labeled samples in the Web and selects the most informative samples for expert labeling such that we can construct effective classifiers for novel classes with only a few labeled samples; and 3) zero-shot learning which utilizes no labeled samples for novel classes at all to build supervised classifiers for them by transferring knowledge from the related classes.
Cyber-physical systems (CPS) are intended to receive information from the environment through sensors and perform appropriate actions using actuators of the controller. In the last years world of intelligent technologies has grown in an exponential fashion: from cruise control to smart ecosystems. Next we are facing the future of CPS involved in almost every aspect of our lives bringing higher comfortability and efficiency. Our goal is to help smart inventions adjust to this highly uncertain environment and guarantee safety for its inhabitants. The physical environment renders the problem of CPS verification extremely cumbersome. Due to a wealth of uncertainties introduced by physical processes, the system is best described by stochastic models. Approximate prediction techniques, such as Statistical Model Checking (SMC), have therefore recently become increasingly popular. As a result, verification of a CPS boils down to quantitative analysis of how close the system is to reaching bad states (safety property) or desired goal (liveness property). Controlling the systems, that is, computing appropriate response actions depending on the environment, involves probabilistic state estimation, as well as optimal action prediction, i.e., choosing the best next step by simulating the future. In my thesis, I develop a novel intelligent algorithm addressing existing deficiencies of SMC such as poor prediction of rare events (RE) and sampling divergence.
Reinforcement Learning (RL) is a widely used solution for sequential decision-making problems and has been used in many complex domains. However, RL algorithms suffer from scalability issues, especially when multiple agents are acting in a shared environment. This research intends to accelerate learning in multiagent sequential decision-making tasks by reusing previous knowledge, both from past solutions and advising between agents. We intend to contribute a Transfer Learning framework focused on Multiagent RL, requiring as few domain-specific hand-coded parameters as possible.
Image Understanding is fundamental to intelligent agents.Researchers have explored Caption Generation and VisualQuestion Answering as independent aspects of Image Understanding (Johnson et al. 2015; Xiong, Merity, and Socher2016). Common to most of the successful approaches, are the learning of end-to-end signal mapping (image-to-caption, image and question to answer). The accuracy is impressive. It is also important to explain a decision to end-user(justify the results, and rectify based on feedback). Very recently, there has been some focus (Hendricks et al. 2016;Liu et al. ) on explaining some aspects of the learning systems. In my research, I look towards building explainableImage Understanding systems that can be used to generate captions and answer questions. Humans learn both from examples (learning) and by reading (knowledge). Inspired by such an intuition, researchers have constructed Knowledge-Bases that encode (probabilistic) commonsense and background knowledge. In this work, we look towards efficiently using this probabilistic knowledge on top of machine learning capabilities, to rectify noise in visual detections and generate captions or answers to posed questions.
Branching story games have gained popularity for adapting to user actions within a story world. An active area of Interactive Narrative (IN) research uses automated planning to generate story plans as it can lighten the authorial burden of writing a branching story. Branches can be generated from a declarative representation rather than hand-crafted. A goal of an Experience Manager (EM) is to guide a user through a space of desirable narrative trajectories, or story branches, in an IN. However, in the cases when an EM must accommodate user actions and mediate them from a desired narrative trajectory to a new narrative trajectory, automated planning’s authorial advantage becomes a liability as the available narrative trajectories are not known apriori. This limitation can lead to the EM choosing a new narrative trajectory that is not coherent with the previous one and may result in a negative user experience. The goal of my research is to develop a problem formulation methodology for story planning problems that elicits the available narrative trajectories enabling an EM to execute more coherent accommodations.
With the rise of the Internet and social media, social mobilization - large-scale mobilization manpower for scientific, social, and political activities through crowdsourcing - has become a widespread practice. Despite the success, social mobilization is not without its limitations. Local trapping of diffusion and the dependence on highly connected individuals to mobilize people in distance locations affect the effectiveness of social mobilization. Furthermore, as empirical studies on people's responses to various social mobilization approaches are lacking, it is a significant challenge for artificial intelligence (AI) researchers to design effective and efficient decision support mechanisms to help manage this emerging phenomenon. In my thesis, I conduct large-scale empirical studies to help the AI research community establish baseline personal variabilities in different people's response patterns to social mobilization approaches. Based on the collected dataset, I will further propose computational algorithmic crowdsourcing mechanisms which leverage the empirical evidence to improve the effectiveness and efficiency of social mobilization, towards achieving superlinear productivity. Throughout this process, I will also incorporate human factors into the computational models to benefit social mobilization efforts.
Multiway data, described by tensors, are common in real-world applications. For example, online advertising click logs can be represented by a three-mode tensor (user, advertisement, context). The analysis of tensors is closely related to many important applications, such as click-through-rate (CTR) prediction, anomaly detection and product recommendation. Despite the success of existing tensor analysis approaches, such as Tucker, CANDECOMP/PARAFAC and infinite Tucker decompositions, they are either not enough powerful to capture complex hidden relationships in data, or not scalable to handle real-world large data. In addition, they may suffer from the extreme sparsity in real data, i.e., when the portion of nonzero entries is extremely low; they lack of principled ways to discover other patterns — such as an unknown number of latent clusters — which are critical for data mining tasks such as anomaly detection and market targeting. To address these challenges, I used nonparametric Bayesian techniques, such as Gaussian processes (GP) and Dirichlet processes (DP), to model highly nonlinear interactions and to extract hidden patterns in tensors; I derived tractable variational evidence lower bounds, based on which I developed scalable, distributed or online approximate inference algorithms. Experiments on both simulation and real-world large data have demonstrated the effect of my propoaed approaches.
Systems deployed in unstructured environments must be able to adapt to novel situations. This requires the ability to perform in domains that may be vastly different from training domains. My dissertation focuses on the representations used in lifelong learning and how these representations enable predictions and knowledge sharing over time, allowing an agent to continuously learn and adapt in changing environments. Specifically, my contributions will enable lifelong learning systems to efficiently accumulate data, use prior knowledge to predict models for novel tasks, and alter existing models to account for changes in the environment.
The web has become the central platform and marketplace for the organization, propagation of events and sale of tickets of any kind. Such events range from concerts, workshops, sport events, professional events to small local events. Single event pages are typically split into a textual event description and a set of core event attributes that are specifically highlighted and presented in the same template for all events of a particular source. In this work, we aim to learn a joint model for the extraction of event attributes from both event descriptions and templates. We also investigate the automatic discovery of event sources and the identification of single event pages within event sources. By considering all three problems as part of an integral system, we can exploit mutual reinforcement between the models derived for each sub problem.
Time series data is common in a wide range of disciplines including finance, biology, sociology, and computer science. Analyzing and modeling time series data is fundamental for studying various problems in those fields. For instance, studying time series physiological data can be used to discriminate patients’ abnormal recovery trajectories and normal ones (Hripcsak, Albers, and Perotte 2015). GPS data are useful for studying collective decision making of groupliving animals (Strandburg-Peshkin et al. 2015). There are different methods for studying time series data such as clustering, regression, and anomaly detection. In this proposal, we are interested in structured prediction problems in time series data. Structured prediction focuses on prediction task where the outputs are structured and interdependent, contrary to the non-structured prediction which assumes that the outputs are independent of other predicted outputs. Structured prediction is an important problem as there are structures inherently existing in time series data. One difficulty for structured prediction is that the number of possible outputs can be exponential which makes modeling all the potential outputs intractable.
Project scheduling is a common business management task. However, current business management environment has become more open and dynamic, which jeopardizes the effectiveness of the traditional approaches. In this abstract, I summarize my works in addressing two variations of project scheduling problems, including a combinatorial auction based approach for solving the decentralized multi-project scheduling problem, and a sampling based approach for solving the problem of project scheduling under time-dependent duration uncertainties.
Multi-Robot Task Allocation (MRTA) has no formal framework which could provide solutions covering different domains within the MRTA taxonomy without changing the optimization scheme. This research aims to develop a novel framework using evolutionary computing. The study proposes a modular approach towards developing this framework in which individual problem types of the MRTA taxonomy are solved one at a time. The performance of the framework will be evaluated against the popular approaches suggested for each problem type.
Learning fast and efficiently using minimal data has been consistently a challenge in machine learning. In my thesis, I explore this problem for knowledge transfer for multi-agent multi-task learning in a life-long learning paradigm. My goal is to demonstrate that by sharing knowledge between agents and similar tasks, efficient algorithms can be designed that can increase the speed of learning as well as improve performance. Moreover, this would allow for handling hard tasks through collective learning of multiple agents that share knowledge. As an initial step, I study the problem of incorporating task descriptors into lifelong learning of related tasks to perform zero-shot knowledge transfer. Zero-shot learning is highly desirable because it leads to considerable speedup in handling similar sequential tasks. Then I focus on a multi-agent learning setting, where related tasks are learned collectively and/or address privacy concerns.
Recent successes in applying Deep Learning techniques on Reinforcement Learning algorithms have led to a wave of breakthrough developments in agent theory and established the field of Deep Reinforcement Learning (DRL). While DRL has shown great results for single task learning, the multi-task case is still underrepresented in the available literature. This D.Sc. research proposal aims at extending DRL to the multi- task case by leveraging the power of Transfer Learning algorithms to improve the training time and results for multi-task learning. Our focus lies on defining a novel framework for scalable DRL agents that detects similarities between tasks and balances various TL techniques, like parameter initialization, policy or skill transfer.
My work on the PhD thesis concerns human-like reasoning about relations between spatial objects and the way they change in time. In particular, my research is focused on logic-based reasoning systems that model human spatial reasoning methods and may enable better understanding of humans reasoning mechanisms in future. Importantly, such formalisms are also interested from the practical point of view – they have a number of potential applications, e.g., in robotics, architecture design, databases, among others.
Electroencephalography (EEG) is one of the most important noninvasive neuroimaging tools that provides excellent temporal accuracy. As the EEG electrode sensors measure electrical potentials on the scalp instead of direct measuring activities of brain voxels deep inside the head, many approaches are proposed to infer the activated brain regions due to its significance in neuroscience research and clinical application. However, since mostly part of the brain activity is composed of the spontaneous neural activities or non-task related activations, task related activation patterns will be corrupted in strong background signal/noises. In our research, we proposed a sparse learning framework for solving EEG inverse problem which aims to explicitly extract the discriminative sources for different cognitive tasks by fusing the label information into the inverse model. The proposed framework is capable of estimation the discriminative brain sources under given different brain states where traditional inverse methods failed. We introduced two models, one is formulated as supervised sparse dictionary learning and the other one is the graph regularized discriminative source estimation model to promote the consistency within same class. Preliminary experimental results also validated that the proposed sparse learning framework is effective to discover the discriminative task-related brain activation sources, which shows the potential to advance the high resolution EEG source analysis for real-time non-invasive brain imaging research.