Learning to act in uncertain environments : technical perspective

The problem of decision making in an uncertain environment arises in many diverse contexts: deciding whether to keep a hard drive spinning in a net-book; choosing which advertisement to post to a Web site visitor; choosing how many newspapers to order so as to maximize profits; or choosing a route to recommend to a driver given limited and possibly out-of-date information about traffic conditions. All are sequential decision problems, since earlier decisions affect subsequent performance; all require adaptive approaches, since they involve significant uncertainty. The key issue in effectively solving problems like these is known as the exploration/exploitation trade-off: If I am at a cross-roads, when should I go in the most advantageous direction among those that I have already explored, and when should I strike out in a new direction, in the hopes I will discover something better?

Exponentiated gradient algorithms for conditional random fields and max-margin Markov Networks

Log-linear and maximum-margin models are two commonly-used methods in supervised machine learning, and are frequently used in structured prediction problems. Efficient learning of parameters in these models is therefore an important problem, and becomes a key factor when learning from very large data sets. This paper describes exponentiated gradient (EG) algorithms for training such models, where EG updates are applied to the convex dual of either the log-linear or max-margin objective function; the dual in both the log-linear and max-margin cases corresponds to minimizing a convex function with simplex constraints. We study both batch and online variants of the algorithm, and provide rates of convergence for both cases. In the max-margin case, O(1/ε) EG updates are required to reach a given accuracy ε in the dual; in contrast, for log-linear models only O(log(1/ε)) updates are required. For both the max-margin and log-linear cases, our bounds suggest that the online EG algorithm requires a factor of n less computation to reach a desired accuracy than the batch EG algorithm, where n is the number of training examples. Our experiments confirm that the online algorithms are much faster than the batch algorithms in practice. We describe how the EG updates factor in a convenient way for structured prediction problems, allowing the algorithms to be efficiently applied to problems such as sequence learning or natural language parsing. We perform extensive evaluation of the algorithms, comparing them to L-BFGS and stochastic gradient descent for log-linear models, and to SVM-Struct for max-margin models. The algorithms are applied to a multi-class problem as well as to a more complex large-scale parsing task. In all these settings, the EG algorithms presented here outperform the other methods.

A unifying view of multiple kernel learning

Recent research on multiple kernel learning has lead to a number of approaches for combining kernels in regularized risk minimization. The proposed approaches include different formulations of objectives and varying regularization strategies. In this paper we present a unifying general optimization criterion for multiple kernel learning and show how existing formulations are subsumed as special cases. We also derive the criterion's dual representation, which is suitable for general smooth optimization algorithms. Finally, we evaluate multiple kernel learning in this framework analytically using a Rademacher complexity bound on the generalization error and empirically in a set of experiments.

Acquiring search expertise : learning experiences and threshold concepts

Expert searchers engage with information as information brokers, researchers, reference librarians, information architects, faculty who teach advanced search, and in a variety of other information-intensive professions. Their experiences are characterized by a profound understanding of information concepts and skills and they have an agile ability to apply this knowledge to interacting with and having an impact on the information environment. This study explored the learning experiences of searchers to understand the acquisition of search expertise. The research question was: What can be learned about becoming an expert searcher from the learning experiences of proficient novice searchers and highly experienced searchers? The key objectives were: (1) to explore the existence of threshold concepts in search expertise; (2) to improve our understanding of how search expertise is acquired and how novice searchers, intent on becoming experts, can learn to search in more expertlike ways. The participant sample drew from two population groups: (1) highly experienced searchers with a minimum of 20 years of relevant professional experience, including LIS faculty who teach advanced search, information brokers, and search engine developers (11 subjects); and (2) MLIS students who had completed coursework in information retrieval and online searching and demonstrated exceptional ability (9 subjects). Using these two groups allowed a nuanced understanding of the experience of learning to search in expertlike ways, with data from those who search at a very high level as well as those who may be actively developing expertise. The study used semi-structured interviews, search tasks with think-aloud narratives, and talk-after protocols. Searches were screen-captured with simultaneous audio-recording of the think-aloud narrative. Data were coded and analyzed using NVivo9 and manually. Grounded theory allowed categories and themes to emerge from the data. Categories represented conceptual knowledge and attributes of expert searchers. In accord with grounded theory method, once theoretical saturation was achieved, during the final stage of analysis the data were viewed through lenses of existing theoretical frameworks. For this study, threshold concept theory (Meyer & Land, 2003) was used to explore which concepts might be threshold concepts. Threshold concepts have been used to explore transformative learning portals in subjects ranging from economics to mathematics. A threshold concept has five defining characteristics: transformative (causing a shift in perception), irreversible (unlikely to be forgotten), integrative (unifying separate concepts), troublesome (initially counter-intuitive), and may be bounded. Themes that emerged provided evidence of four concepts which had the characteristics of threshold concepts. These were: information environment: the total information environment is perceived and understood; information structures: content, index structures, and retrieval algorithms are understood; information vocabularies: fluency in search behaviors related to language, including natural language, controlled vocabulary, and finesse using proximity, truncation, and other language-based tools. The fourth threshold concept was concept fusion, the integration of the other three threshold concepts and further defined by three properties: visioning (anticipating next moves), being light on one's 'search feet' (dancing property), and profound ontological shift (identity as searcher). In addition to the threshold concepts, findings were reported that were not concept-based, including praxes and traits of expert searchers. A model of search expertise is proposed with the four threshold concepts at its core that also integrates the traits and praxes elicited from the study, attributes which are likewise long recognized in LIS research as present in professional searchers. The research provides a deeper understanding of the transformative learning experiences involved in the acquisition of search expertise. It adds to our understanding of search expertise in the context of today's information environment and has implications for teaching advanced search, for research more broadly within library and information science, and for methodologies used to explore threshold concepts.

Document clustering algorithms, representations and evaluation for information retrieval

This thesis presents new methods for classification and thematic grouping of billions of web pages, at scales previously not achievable. This process is also known as document clustering, where similar documents are automatically associated with clusters that represent various distinct topic. These automatically discovered topics are in turn used to improve search engine performance by only searching the topics that are deemed relevant to particular user queries.

Learning-based Relevance Feedback for Web-based Relation Completion

In a pilot application based on web search engine calledWeb-based Relation Completion (WebRC), we propose to join two columns of entities linked by a predefined relation by mining knowledge from the web through a web search engine. To achieve this, a novel retrieval task Relation Query Expansion (RelQE) is modelled: given an entity (query), the task is to retrieve documents containing entities in predefined relation to the given one. Solving this problem entails expanding the query before submitting it to a web search engine to ensure that mostly documents containing the linked entity are returned in the top K search results. In this paper, we propose a novel Learning-based Relevance Feedback (LRF) approach to solve this retrieval task. Expansion terms are learned from training pairs of entities linked by the predefined relation and applied to new entity-queries to find entities linked by the same relation. After describing the approach, we present experimental results on real-world web data collections, which show that the LRF approach always improves the precision of top-ranked search results to up to 8.6 times the baseline. Using LRF, WebRC also shows performances way above the baseline.

Machine learning for activity recognition : hip versus wrist data

Problem addressed Wrist-worn accelerometers are associated with greater compliance. However, validated algorithms for predicting activity type from wrist-worn accelerometer data are lacking. This study compared the activity recognition rates of an activity classifier trained on acceleration signal collected on the wrist and hip. Methodology 52 children and adolescents (mean age 13.7 +/- 3.1 year) completed 12 activity trials that were categorized into 7 activity classes: lying down, sitting, standing, walking, running, basketball, and dancing. During each trial, participants wore an ActiGraph GT3X+ tri-axial accelerometer on the right hip and the non-dominant wrist. Features were extracted from 10-s windows and inputted into a regularized logistic regression model using R (Glmnet + L1). Results Classification accuracy for the hip and wrist was 91.0% +/- 3.1% and 88.4% +/- 3.0%, respectively. The hip model exhibited excellent classification accuracy for sitting (91.3%), standing (95.8%), walking (95.8%), and running (96.8%); acceptable classification accuracy for lying down (88.3%) and basketball (81.9%); and modest accuracy for dance (64.1%). The wrist model exhibited excellent classification accuracy for sitting (93.0%), standing (91.7%), and walking (95.8%); acceptable classification accuracy for basketball (86.0%); and modest accuracy for running (78.8%), lying down (74.6%) and dance (69.4%). Potential Impact Both the hip and wrist algorithms achieved acceptable classification accuracy, allowing researchers to use either placement for activity recognition.

Selection procedures for module discovery : exploring evolutionary algorithms for cognitive science

Evolutionary algorithms are playing an increasingly important role as search methods in cognitive science domains. In this study, methodological issues in the use of evolutionary algorithms were investigated via simulations in which procedures were systematically varied to modify the selection pressures on populations of evolving agents. Traditional roulette wheel, tournament, and variations of these selection algorithms were compared on the “needle-in-a-haystack” problem developed by Hinton and Nowlan in their 1987 study of the Baldwin effect. The task is an important one for cognitive science, as it demonstrates the power of learning as a local search technique in smoothing a fitness landscape that lacks gradient information. One aspect that has continued to foster interest in the problem is the observation of residual learning ability in simulated populations even after long periods of time. Effective evolutionary algorithms balance their search effort between broad exploration of the search space and in-depth exploitation of promising solutions already found. Issues discussed include the differential effects of rank and proportional selection, the tradeoff between migration of populations towards good solutions and maintenance of diversity, and the development of measures that illustrate how each selection algorithm affects the search process over generations. We show that both roulette wheel and tournament algorithms can be modified to appropriately balance search between exploration and exploitation, and effectively eliminate residual learning in this problem.

Learning the learning rate for prediction with expert advice

Most standard algorithms for prediction with expert advice depend on a parameter called the learning rate. This learning rate needs to be large enough to fit the data well, but small enough to prevent overfitting. For the exponential weights algorithm, a sequence of prior work has established theoretical guarantees for higher and higher data-dependent tunings of the learning rate, which allow for increasingly aggressive learning. But in practice such theoretical tunings often still perform worse (as measured by their regret) than ad hoc tuning with an even higher learning rate. To close the gap between theory and practice we introduce an approach to learn the learning rate. Up to a factor that is at most (poly)logarithmic in the number of experts and the inverse of the learning rate, our method performs as well as if we would know the empirically best learning rate from a large range that includes both conservative small values and values that are much higher than those for which formal guarantees were previously available. Our method employs a grid of learning rates, yet runs in linear time regardless of the size of the grid.

Automating marine mammal detection in aerial images captured during wildlife surveys: A deep learning approach

Aerial surveys conducted using manned or unmanned aircraft with customized camera payloads can generate a large number of images. Manual review of these images to extract data is prohibitive in terms of time and financial resources, thus providing strong incentive to automate this process using computer vision systems. There are potential applications for these automated systems in areas such as surveillance and monitoring, precision agriculture, law enforcement, asset inspection, and wildlife assessment. In this paper, we present an efficient machine learning system for automating the detection of marine species in aerial imagery. The effectiveness of our approach can be credited to the combination of a well-suited region proposal method and the use of Deep Convolutional Neural Networks (DCNNs). In comparison to previous algorithms designed for the same purpose, we have been able to dramatically improve recall to more than 80% and improve precision to 27% by using DCNNs as the core approach.

The logic of data-sense: Thinking through Learning Personalisation

Big Data and Learning Analytics’ promise to revolutionise educational institutions, endeavours, and actions through more and better data is now compelling. Multiple, and continually updating, data sets produce a new sense of ‘personalised learning’. A crucial attribute of the datafication, and subsequent profiling, of learner behaviour and engagement is the continual modification of the learning environment to induce greater levels of investment on the parts of each learner. The assumption is that more and better data, gathered faster and fed into ever-updating algorithms, provide more complete tools to understand, and therefore improve, learning experiences through adaptive personalisation. The argument in this paper is that Learning Personalisation names a new logistics of investment as the common ‘sense’ of the school, in which disciplinary education is ‘both disappearing and giving way to frightful continual training, to continual monitoring'.