Technology has helped to innovate in the teaching-learning process. Today's students are more demanding actors when it comes to the environment, they have at their disposal to learn, experiment and develop critical thinking. The area of mathematics has successively suffered from students' learning difficulties, whether due to lack of motivation, low abstraction ability, or lack of new tools for teachers to bring innovation into the classroom and outside it. While it is true that digitalization has entered schools, it often follows a process of digital replication of approaches and materials that were previously only available on physical media. This work focuses on the use of Extended Realities for teaching mathematics, and very particularly in the teaching of geometry, with a proposition of a conceptual model that combines the use of Extended Reality and Machine Learning. The proposed model was subject to prototyping, which is presented as a form of laboratory validation as a contribution to innovate the way in which the geometry teaching-learning process is developed, as well as through the ability to obtain useful insights for teachers and students throughout the process.
Online active learning is a paradigm in machine learning that aims to select the most informative data points to label from a data stream. The problem of minimizing the cost associated with collecting labeled observations has gained a lot of attention in recent years, particularly in real-world applications where data is only available in an unlabeled form. Annotating each observation can be time-consuming and costly, making it difficult to obtain large amounts of labeled data. To overcome this issue, many active learning strategies have been proposed in the last decades, aiming to select the most informative observations for labeling in order to improve the performance of machine learning models. These approaches can be broadly divided into two categories: static pool-based and stream-based active learning. Pool-based active learning involves selecting a subset of observations from a closed pool of unlabeled data, and it has been the focus of many surveys and literature reviews. However, the growing availability of data streams has led to an increase in the number of approaches that focus on online active learning, which involves continuously selecting and labeling observations as
With the increasing interest in deploying Artificial Intelligence in medicine, we previously introduced HAIM (Holistic AI in Medicine), a framework that fuses multimodal data to solve downstream clinical tasks. However, HAIM uses data in a task-agnostic manner and lacks explainability. To address these limitations, we introduce xHAIM (Explainable HAIM), a novel framework leveraging Generative AI to enhance both prediction and explainability through four structured steps: (1) automatically identifying task-relevant patient data across modalities, (2) generating comprehensive patient summaries, (3) using these summaries for improved predictive modeling, and (4) providing clinical explanations by linking predictions to patient-specific medical knowledge. Evaluated on the HAIM-MIMIC-MM dataset, xHAIM improves average AUC from 79.9% to 90.3% across chest pathology and operative tasks. Importantly, xHAIM transforms AI from a black-box predictor into an explainable decision support system, enabling clinicians to interactively trace predictions back to relevant patient data, bridging AI advancements with clinical utility.
A key challenge for machine intelligence is to learn new visual concepts without forgetting the previously acquired knowledge. Continual learning is aimed towards addressing this challenge. However, there is a gap between existing supervised continual learning and human-like intelligence, where human is able to learn from both labeled and unlabeled data. How unlabeled data affects learning and catastrophic forgetting in the continual learning process remains unknown. To explore these issues, we formulate a new semi-supervised continual learning method, which can be generically applied to existing continual learning models. Specifically, a novel gradient learner learns from labeled data to predict gradients on unlabeled data. Hence, the unlabeled data could fit into the supervised continual learning method. Different from conventional semi-supervised settings, we do not hypothesize that the underlying classes, which are associated to the unlabeled data, are known to the learning process. In other words, the unlabeled data could be very distinct from the labeled data. We evaluate the proposed method on mainstream continual learning, adversarial continual learning, and semi-supervised
In 1949, Captain Alberto Larraguibel and his horse Huaso set the world record for equestrian high jump in Viña del Mar, Chile, by clearing a height of 2.47 meters, a mark that remains unbeaten. This work proposes the use of this historical event as a teaching resource for physics, integrating perspectives from biomechanics and veterinary medicine. Based on the analysis of an audiovisual record of the jump, a kinematic model is developed using the \textit{Tracker} software, determining variables such as displacement, velocity, and acceleration of the horse--rider system. The results make it possible to reflect on the biomechanical and physiological factors involved in animal performance, thus linking physics with real biological processes. It is proposed that this interdisciplinary approach, based on authentic cultural and scientific contexts, may promote meaningful learning, motivation, and a more comprehensive understanding of natural phenomena in science education.
Recruiting and retaining highly qualified physics and physical science teachers is critical for maintaining America's global competitiveness. Unfortunately, only one third of the high school teachers in physics have a degree in physics and an even smaller number of physical science teachers in middle school have a good grasp of the scientific content they teach. Moreover, teachers often lack adequate pedagogical content knowledge to teach science effectively. Here, we discuss the development, implementation, and assessment of a course for science and engineering undergraduates designed to increase awareness and help them develop an interest and a deeper appreciation of the intellectual demands of physics teaching. The course focused on increasing student enthusiasm and confidence in teaching by providing well supported teaching opportunities and exposure to physics education research. The course assessment methods include 1) pre/post-test measures of attitude and expectations about science teaching, 2) self and peer evaluation of student teaching, 3) content-based pre/post-tests given to students who received instruction from the student teachers, and 4) audio-taped focus group dis
Machine teaching is an inverse problem of machine learning that aims at steering the student learner towards its target hypothesis, in which the teacher has already known the student's learning parameters. Previous studies on machine teaching focused on balancing the teaching risk and cost to find those best teaching examples deriving the student model. This optimization solver is in general ineffective when the student learner does not disclose any cue of the learning parameters. To supervise such a teaching scenario, this paper presents a distribution matching-based machine teaching strategy. Specifically, this strategy backwardly and iteratively performs the halving operation on the teaching cost to find a desired teaching set. Technically, our strategy can be expressed as a cost-controlled optimization process that finds the optimal teaching examples without further exploring in the parameter distribution of the student learner. Then, given any a limited teaching cost, the training examples will be closed-form. Theoretical analysis and experiment results demonstrate this strategy.
Applying Design Science Research (DSR) methodology is becoming a popular working resource for most Information Systems (IS) and Software engineering studies. The research and/or practical design problems that must be faced aim to answer the question of how to create or investigate an artifact in a given context. Precisely characterizing both artifact and context is essential for effective research development. While various design science guidelines and frameworks have been created by experts in IS engineering, emerging researchers and postgraduate students still find it challenging to apply this research methodology correctly. There is limited literature and materials that guide and support teaching novice researchers about the types of artifacts that can be developed to address a particular problem and decision-making in DSR. To address this gap in DSR, in this chapter, we explore DSR from an educational perspective, explaining both the concept of DSR and an effective method for teaching it. This chapter includes examples of DSR, a teaching methodology, learning objectives, and recommendations. Moreover, we have created a survey artifact intended to gather data on the experiences
We study the task of teaching a machine to classify objects using features and labels. We introduce the Error-Driven-Featuring design pattern for teaching using features and labels in which a teacher prefers to introduce features only if they are needed. We analyze the potential risks and benefits of this teaching pattern through the use of teaching protocols, illustrative examples, and by providing bounds on the effort required for an optimal machine teacher using a linear learning algorithm, the most commonly used type of learners in interactive machine learning systems. Our analysis provides a deeper understanding of potential trade-offs of using different learning algorithms and between the effort required for featuring (creating new features) and labeling (providing labels for objects).
In sequential machine teaching, a teacher's objective is to provide the optimal sequence of inputs to sequential learners in order to guide them towards the best model. In this paper we extend this setting from current static one-data-set analyses to learners which change their learning algorithm or latent state to improve during learning, and to generalize to new datasets. We introduce a multi-agent formulation in which learners' inner state may change with the teaching interaction, which affects the learning performance in future tasks. In order to teach such learners, we propose an optimal control approach that takes the future performance of the learner after teaching into account. This provides tools for modelling learners having inner states, and machine teaching of meta-learning algorithms. Furthermore, we distinguish manipulative teaching, which can be done by effectively hiding data and also used for indoctrination, from more general education which aims to help the learner become better at generalization and learning in new datasets in the absence of a teacher.
The topic of comprehensibility of machine-learned theories has recently drawn increasing attention. Inductive Logic Programming (ILP) uses logic programming to derive logic theories from small data based on abduction and induction techniques. Learned theories are represented in the form of rules as declarative descriptions of obtained knowledge. In earlier work, the authors provided the first evidence of a measurable increase in human comprehension based on machine-learned logic rules for simple classification tasks. In a later study, it was found that the presentation of machine-learned explanations to humans can produce both beneficial and harmful effects in the context of game learning. We continue our investigation of comprehensibility by examining the effects of the ordering of concept presentations on human comprehension. In this work, we examine the explanatory effects of curriculum order and the presence of machine-learned explanations for sequential problem-solving. We show that 1) there exist tasks A and B such that learning A before B has a better human comprehension with respect to learning B before A and 2) there exist tasks A and B such that the presence of explanatio
Ethnography has become one of the established methods for empirical research on software engineering. Although there is a wide variety of introductory books available, there has been no material targeting software engineering students particularly, until now. In this chapter we provide an introduction to teaching and learning ethnography for faculty teaching ethnography to software engineering graduate students and for the students themselves of such courses. The contents of the chapter focuses on what we think is the core basic knowledge for newbies to ethnography as a research method. We complement the text with proposals for exercises, tips for teaching, and pitfalls that we and our students have experienced. The chapter is designed to support part of a course on empirical software engineering and provides pointers and literature for further reading.
The last decade has seen an explosion in models that describe phenomena in systems medicine. Such models are especially useful for studying signaling pathways, such as the Wnt pathway. In this chapter we use the Wnt pathway to showcase current mathematical and statistical techniques that enable modelers to gain insight into (models of) gene regulation, and generate testable predictions. We introduce a range of modeling frameworks, but focus on ordinary differential equation (ODE) models since they remain the most widely used approach in systems biology and medicine and continue to offer great potential. We present methods for the analysis of a single model, comprising applications of standard dynamical systems approaches such as nondimensionalization, steady state, asymptotic and sensitivity analysis, and more recent statistical and algebraic approaches to compare models with data. We present parameter estimation and model comparison techniques, focusing on Bayesian analysis and coplanarity via algebraic geometry. Our intention is that this (non exhaustive) review may serve as a useful starting point for the analysis of models in systems medicine.
Contrastive learning operates on a simple yet effective principle: Embeddings of positive pairs are pulled together, while those of negative pairs are pushed apart. In this paper, we propose a unified framework for understanding contrastive learning through the lens of cosine similarity, and present two key theoretical insights derived from this framework. First, in full-batch settings, we show that perfect alignment of positive pairs is unattainable when negative-pair similarities fall below a threshold, and this misalignment can be mitigated by incorporating within-view negative pairs into the objective. Second, in mini-batch settings, smaller batch sizes induce stronger separation among negative pairs in the embedding space, i.e., higher variance in their similarities, which in turn degrades the quality of learned representations compared to full-batch settings. To address this, we propose an auxiliary loss that reduces the variance of negative-pair similarities in mini-batch settings. Empirical results show that incorporating the proposed loss improves performance in small-batch settings.
Data science has become increasingly essential for the production of official statistics, as it enables the automated collection, processing, and analysis of large amounts of data. With such data science practices in place, it enables more timely, more insightful and more flexible reporting. However, the quality and integrity of data-science-driven statistics rely on the accuracy and reliability of the data sources and the machine learning techniques that support them. In particular, changes in data sources are inevitable to occur and pose significant risks that are crucial to address in the context of machine learning for official statistics. This paper gives an overview of the main risks, liabilities, and uncertainties associated with changing data sources in the context of machine learning for official statistics. We provide a checklist of the most prevalent origins and causes of changing data sources; not only on a technical level but also regarding ownership, ethics, regulation, and public perception. Next, we highlight the repercussions of changing data sources on statistical reporting. These include technical effects such as concept drift, bias, availability, validity, accur
Machine teaching addresses the problem of finding the best training data that can guide a learning algorithm to a target model with minimal effort. In conventional settings, a teacher provides data that are consistent with the true data distribution. However, for sequential learners which actively choose their queries, such as multi-armed bandits and active learners, the teacher can only provide responses to the learner's queries, not design the full data. In this setting, consistent teachers can be sub-optimal for finite horizons. We formulate this sequential teaching problem, which current techniques in machine teaching do not address, as a Markov decision process, with the dynamics nesting a model of the learner and the actions being the teacher's responses. Furthermore, we address the complementary problem of learning from a teacher that plans: to recognise the teaching intent of the responses, the learner is endowed with a model of the teacher. We test the formulation with multi-armed bandit learners in simulated experiments and a user study. The results show that learning is improved by (i) planning teaching and (ii) the learner having a model of the teacher. The approach giv
Model-agnostic meta-learning (MAML) is a meta-learning technique to train a model on a multitude of learning tasks in a way that primes the model for few-shot learning of new tasks. The MAML algorithm performs well on few-shot learning problems in classification, regression, and fine-tuning of policy gradients in reinforcement learning, but comes with the need for costly hyperparameter tuning for training stability. We address this shortcoming by introducing an extension to MAML, called Alpha MAML, to incorporate an online hyperparameter adaptation scheme that eliminates the need to tune meta-learning and learning rates. Our results with the Omniglot database demonstrate a substantial reduction in the need to tune MAML training hyperparameters and improvement to training stability with less sensitivity to hyperparameter choice.
We motivate Energy-Based Models (EBMs) as a promising model class for continual learning problems. Instead of tackling continual learning via the use of external memory, growing models, or regularization, EBMs change the underlying training objective to cause less interference with previously learned information. Our proposed version of EBMs for continual learning is simple, efficient, and outperforms baseline methods by a large margin on several benchmarks. Moreover, our proposed contrastive divergence-based training objective can be combined with other continual learning methods, resulting in substantial boosts in their performance. We further show that EBMs are adaptable to a more general continual learning setting where the data distribution changes without the notion of explicitly delineated tasks. These observations point towards EBMs as a useful building block for future continual learning methods.
In this paper we provide an account of how we ported a text and data mining course online in summer 2020 as a result of the COVID-19 pandemic and how we improved it in a second pilot run. We describe the course, how we adapted it over the two pilot runs and what teaching techniques we used to improve students' learning and community building online. We also provide information on the relentless feedback collected during the course which helped us to adapt our teaching from one session to the next and one pilot to the next. We discuss the lessons learned and promote the use of innovative teaching techniques applied to the digital such as digital badges and pair programming in break-out rooms for teaching Natural Language Processing courses to beginners and students with different backgrounds.
In this paper we put forth a model for physics course reform that uniquely uses proven, research-based active learning strategies to help students improve their physics knowledge and problem-solving skills. In this study, we compared the exam performance of students in two sections of the same introductory physics course. One section (the traditional section, N = 258) was taught by an instructor who is highly regarded for his lectures, but did not use any active learning teaching strategies. The other section (the reformed section, N = 217) was taught by an instructor who had never before taught a physics class but who was trained in physics and astronomy education research and who did use active learning teaching strategies. Students in the reformed section significantly outperformed students in the traditional section on common exam questions over the course of the semester, regardless of whether the question was conceptual or quantitative. This reform effort has been successful at improving students' learning and significantly increasing the department's use of active learning strategies at the introductory level and beyond.