Active Learning Setting Research Articles

Active learning in neuroscience: a manipulative to simulate visual field defects.

prevalent in 20–57% of stroke patients, visual field defects have been shown to impact quality of life. Studies ([8][1]) have shown increased risk of falling, ambulatory difficulties, impaired reading ability, and feelings of panic in crowded or unfamiliar places in patients with visual field

Enterprise system case using Microsoft Dynamics GP via DynamicsCloud

Abstract This case increases your understanding of enterprise systems by applying course concepts in an active-learning setting. Specifically, you build on previous course knowledge by using Microsoft Dynamics GP via a cloud provider, DynamicsCloud. Cloud applications allow companies to reduce information technology costs and use one consistent platform across the entire company, whether they have one or many locations across the globe. The case consists of three main parts (i.e., system overview, data collection and storage, and information retrieval) and integrates internal controls and exercises within each part of the case. Upon completion of this case, you will understand the basic functions of an enterprise system. Activities require you to collect and process data in the revenue and purchasing cycles, identify internal controls within the system, and provide useful information for decision making. In addition, you gain hands-on experience using an enterprise system and acquire skills that transfer to other enterprise systems or to upgraded versions of Dynamics GP, which companies can use to support global operations. The case should increase accounting information system knowledge, system navigation skills, and the ability to learn on your own.

Online Passive-Aggressive Active learning

We investigate online active learning techniques for online classification tasks. Unlike traditional supervised learning approaches, either batch or online learning, which often require to request class labels of each incoming instance, online active learning queries only a subset of informative incoming instances to update the classification model, aiming to maximize classification performance with minimal human labelling effort during the entire online learning task. In this paper, we present a new family of online active learning algorithms called Passive-Aggressive Active (PAA) learning algorithms by adapting the Passive-Aggressive algorithms in online active learning settings. Unlike conventional Perceptron-based approaches that employ only the misclassified instances for updating the model, the proposed PAA learning algorithms not only use the misclassified instances to update the classifier, but also exploit correctly classified examples with low prediction confidence. Specifically, we propose several variants of PAA algorithms to tackle three types of online learning tasks: binary classification, multi-class classification, and cost-sensitive classification. We give the mistake bounds of the proposed algorithms in theory, and conduct extensive experiments to evaluate the empirical performance of our techniques on both standard and large-scale datasets, in which the encouraging results validate the empirical effectiveness of the proposed algorithms.

Motivation, Ownership, and the Role of the Instructor in Active Learning

This article systematizes recently published accounts and the author’s experience on motivation, learning process ownership, and the role of the instructor in active learning methods, such as role-plays. It argues that active learning is most productive when the instructor creates and consistently enforces a stable and clear framework of rules. Active learning methods often produce higher levels of motivation, thanks to a greater visibility of students’ efforts, higher stakes of success within a competitive collaborative setup, and a greater ownership of the whole process by students. High levels of motivation arguably improve the quality and depth of learning. Novice users of active learning may, therefore, perceive that the self-effacement of the instructor in a fully horizontal cooperative setting is the right way to optimize learning outcomes. However, ownership and increased motivation are likely to create several problems—overmotivation, confusion, alienation, and hostility—that can be solved best by the instructor. This article uses a conceptualization of the student–instructor relationship in which cooperation and dominance are not mutually exclusive and claims that instead of self-effacement, a cooperative, but authoritative, relationship between students and instructor maximizes the learning potential in active learning settings..

Cross-Modal Similarity Learning via Pairs, Preferences, and Active Supervision

We present a probabilistic framework for learning pairwise similarities between objects belonging to different modalities, such as drugs and proteins, or text and images. Our framework is based on learning a binary code based representation for objects in each modality, and has the following key properties: (i) it can leverage both pairwise as well as easy-to-obtain relative preference based cross-modal constraints, (ii) the probabilistic framework naturally allows querying for the most useful/informative constraints, facilitating an active learning setting (existing methods for cross-modal similarity learning do not have such a mechanism), and (iii) the binary code length is learned from the data. We demonstrate the effectiveness of the proposed approach on two problems that require computing pairwise similarities between cross-modal object pairs: cross-modal link prediction in bipartite graphs, and hashing based cross-modal similarity search.

Querying Discriminative and Representative Samples for Batch Mode Active Learning

Empirical risk minimization (ERM) provides a principled guideline for many machine learning and data mining algorithms. Under the ERM principle, one minimizes an upper bound of the true risk, which is approximated by the summation of empirical risk and the complexity of the candidate classifier class. To guarantee a satisfactory learning performance, ERM requires that the training data are i.i.d. sampled from the unknown source distribution. However, this may not be the case in active learning, where one selects the most informative samples to label, and these data may not follow the source distribution. In this article, we generalize the ERM principle to the active learning setting. We derive a novel form of upper bound for the true risk in the active learning setting; by minimizing this upper bound, we develop a practical batch mode active learning method. The proposed formulation involves a nonconvex integer programming optimization problem. We solve it efficiently by an alternating optimization method. Our method is shown to query the most informative samples while preserving the source distribution as much as possible, thus identifying the most uncertain and representative queries. We further extend our method to multiclass active learning by introducing novel pseudolabels in the multiclass case and developing an efficient algorithm. Experiments on benchmark datasets and real-world applications demonstrate the superior performance of our proposed method compared to state-of-the-art methods.

Optimizing Instance Selection for Statistical Machine Translation with Feature Decay Algorithms

We introduce FDA5 for efficient parameterization, optimization, and implementation of feature decay algorithms (FDA), a class of instance selection algorithms that use feature decay. FDA increase the diversity of the selected training set by devaluing features (i.e., n-grams) that have already been included. FDA5 decides which instances to select based on three functions used for initializing and decaying feature values and scaling sentence scores controlled with five parameters. We present optimization techniques that allow FDA5 to adapt these functions to in-domain and out-of-domain translation tasks for different language pairs. In a transductive learning setting, selection of training instances relevant to the test set can improve the final translation quality. In machine translation experiments performed on the 2 million sentence English-German section of the Europarl corpus, we show that a subset of the training set selected by FDA5 can gain up to 3.22 BLEU points compared to a randomly selected subset of the same size, can gain up to 0.41 BLEU points compared to using all of the available training data using only 15% of it, and can reach within 0.5 BLEU points to the full training set result by using only 2.7% of the full training data. FDA5 peaks at around 8M words or 15% of the full training set. In an active learning setting, FDA5 minimizes the human effort by identifying the most informative sentences for translation and FDA gains up to 0.45 BLEU points using 3/5 of the available training data compared to using all of it and 1.12 BLEU points compared to random training set. In translation tasks involving English and Turkish, a morphologically rich language, FDA5 can gain up to 11.52 BLEU points compared to a randomly selected subset of the same size, can achieve the same BLEU score using as little as 4% of the data compared to random instance selection, and can exceed the full dataset result by 0.78 BLEU points. FDA5 is able to reduce the time to build a statistical machine translation system to about half with 1M words using only 3% of the space for the phrase table and 8% of the overall space when compared with a baseline system using all of the training data available yet still obtain only 0.58 BLEU points difference with the baseline system in out-of-domain translation.

Bayesian Active Remote Sensing Image Classification

In recent years, kernel methods, in particular support vector machines (SVMs), have been successfully introduced to remote sensing image classification. Their properties make them appropriate for dealing with a high number of image features and a low number of available labeled spectra. The introduction of alternative approaches based on (parametric) Bayesian inference has been quite scarce in the more recent years. Assuming a particular prior data distribution may lead to poor results in remote sensing problems because of the specificities and complexity of the data. In this context, the emerging field of nonparametric Bayesian methods constitutes a proper theoretical framework to tackle the remote sensing image classification problem. This paper exploits the Bayesian modeling and inference paradigm to tackle the problem of kernel-based remote sensing image classification. This Bayesian methodology is appropriate for both finite- and infinite-dimensional feature spaces. The particular problem of active learning is addressed by proposing an incremental/active learning approach based on three different approaches: 1) the maximum differential of entropies; 2) the minimum distance to decision boundary; and 3) the minimum normalized distance. Parameters are estimated by using the evidence Bayesian approach, the kernel trick, and the marginal distribution of the observations instead of the posterior distribution of the adaptive parameters. This approach allows us to deal with infinite-dimensional feature spaces. The proposed approach is tested on the challenging problem of urban monitoring from multispectral and synthetic aperture radar data and in multiclass land cover classification of hyperspectral images, in both purely supervised and active learning settings. Similar results are obtained when compared to SVMs in the supervised mode, with the advantage of providing posterior estimates for classification and automatic parameter learning. Comparison with random sampling as well as standard active learning methods such as margin sampling and entropy-query-by-bagging reveals a systematic overall accuracy gain and faster convergence with the number of queries.

An examination of the effects of argument mapping on students’ memory and comprehension performance

Argument mapping (AM) is a method of visually diagramming arguments to allow for easy comprehension of core statements and relations. A series of three experiments compared argument map reading and construction with hierarchical outlining, text summarisation, and text reading as learning methods by examining subsequent memory and comprehension performance. Effects of study environment, argument size, learning strategy (active and passive) and recall interval (immediate and delayed) were also examined. Results revealed that argument map reading and construction significantly increased subsequent immediate recall for arguments in both passive and active learning settings. These findings indicate that AM is a useful learning and teaching methodology, particularly in comparison with standard text-based learning. Results are discussed in light of research and theory on learning and memory.

Approaches to Community Flood Science Engagement: The River Severn Catchment, UK as Case-study

Flood education forms a key element in preparedness strategies to increase flood risk awareness, and empower floodplain communities to act to manage personal risk. Flood science is variably part of education programmes, and yet has potential to increase communities’ capacity to make informed judgements about local risk, and potential mitigation options. The paper draws on action research from two community science engagement projects facilitated by the author in the lower River Severn catchment, UK. Five key questions around community flood science education are explored: 1. Why should the public in community settings engage with flood science? 2. How do the nature of floods, flood risk and flood science impact on approaches to engagement for learning? 3. What should be the relationship between expert flood science and informal science in establishing what should be learnt? 4. What are the key questions for learning in community flood science? 5. What pedagogic approaches to flood science engagement work for the public in community settings? The paper then examines different active learning approaches and settings for community flood science education, and positions these within broader ‘science and society’ contexts for capacity building and empowering communities in the face of changing flood risk.

Analysis of Perceptron-Based Active Learning

We start by showing that in an active learning setting, the Perceptron algorithm needs Ω(1/e2) labels to learn linear separators within generalization error e. We then present a simple active learn...

Don't Throw Out the Baby With the Bathwater

The effectiveness of using verbal repetition and first-letter acronyms to teach a common marketing framework was examined in two experiments. In Experiment 1, 345 undergraduate students were exposed to the framework using one of four conditions: control, verbal repetition, acronym, and verbal repetition plus acronym in a traditional learning setting. Students were tested for unaided recall of the concepts as well as concept application and analysis. Results indicate that using acronyms increased student scores at 2 weeks and 3 months for both unaided recall and analysis, but verbal repetition had no significant effect, either alone or in conjunction with the acronym. Experiment 2 tested the impact of acronym use in an active learning setting. Here, 129 undergraduate students were exposed to the framework using only an active learning method or the active learning method plus an acronym. Students were tested for unaided recall and concept application and analysis at 2 weeks and 3 months after exposure. Use of the acronym increased scores for both unaided recall and concept application and analysis compared to the active learning method alone. Implications for teaching strategies are discussed.

Crop Management Teaching Old Dogs Survival Tricks: A Case Study in Promoting Integrated Crop Management

New technology in crop production is emerging that can provide site-specific field and crop detail information to some operators. Satellite positioning, commonly referred to as global positioning systems, mapping software, and yield-monitoring equipment are all high-tech methods of providing this precision agriculture. But precision agriculture, or at least long strides in that direction, is possible short of these methods and capital investments. Integrated Crop Management (ICM) is one alternative to providing information-intensive management and precision farming, without the need for hard technology. Major elements of an ICM program include thorough crop production and protection planning, crop rotations, tillage management, and nutrient testing, as well as varying fertility rates based on manure and legume credits, scouting for weeds/insects/diseases, and keeping and analyzing field based records. Inherent in ICM is the involvement of the producer in increasing management tools and the knowledge base to make a decision. This knowledge base is generated to a significant degree from on-farm or site-specific observations. The Iowa Model Farms Demonstration Project (IMFDP) was an extension service program established to encourage farmers to adopt ICM practices. Evaluation of the program reveals that opportunities existed for refined practices among the sample of cooperators, although this was not recognized by all, as nearly half of the original participants did not complete the 3-yr project. Among those who remained in the project, several practices were refined and resulted in increased profitability. Among participants who discontinued the program, many perceived the graduated user fee was not justified from the benefits they experienced. Project staff acknowledged they did a poor job of stressing the successes and noting cost savings made by these producers. Lack of detailed records did not allow a clear analysis of the cost-benefit ratio of the program. While documented success of ICM is necessary, a relationship of trust between the consultant and producer is also essential, for successful implementation of ICM practices.