The importance of retention rate for higher education institutions has encouraged data analysts to present various methods to predict at-risk students. The present study, motivated by the same encouragement, proposes a deep learning model trained with 121 features of diverse categories extracted or engineered out of the records of 60,822 postsecondary students. The model undertakes to identify students likely to graduate, the ones likely to transfer to a different school, and the ones likely to drop out and leave their higher education unfinished. This study undertakes to adjust its predictive methods for different stages of curricular progress of students. The temporal aspects introduced for this purpose are accounted for by incorporating layers of LSTM in the model. Our experiments demonstrate that distinguishing between to-be-graduate and at-risk students is reasonably achievable in the earliest stages, and then it rapidly improves, but the resolution within the latter category (dropout vs. transfer) depends on data accumulated over time. However, the model remarkably foresees the fate of students who stay in the school for three years. The model is also assigned to present the weightiest features in the procedure of prediction, both on institutional and student levels. A large, diverse sample size along with the investigation of more than one hundred extracted or engineered features in our study provide new insights into variables that affect students success, predict dropouts with reasonable accuracy, and shed light on the less investigated issue of transfer between colleges. More importantly, by providing individual-level predictions (as opposed to school-level predictions) and addressing the outcomes of transfers, this study improves the use of ML in the prediction of educational outcomes.

This paper proposes a model to predict the levels (e.g., Bachelor, Master, etc.) of postsecondary degree awards that have been ambiguously expressed in the student tracking reports of the National Student Clearinghouse (NSC). The model will be the hybrid of two modules. The first module interprets the relevant abbreviatory elements embedded in NSC reports by referring to a comprehensive database that we have made of nearly 950 abbreviations for degree titles used by American postsecondary educators. The second module is a combination of feature classification and text mining modeled with CNN-BiLSTM, which is preceded by several steps of heavy pre-processing. The model proposed in this paper was trained with four multi-label datasets of different grades of resolution and returned 97.83\% accuracy with the most sophisticated dataset. Such a thorough classification of degree levels will provide insights into the modeling patterns of student success and mobility. To date, such a classification strategy has not been attempted except using manual methods and simple text parsing logic.

Large language models (LLMs) are routinely pre-trained on billions of tokens, only to restart the process over again once new data becomes available. A much cheaper and more efficient solution would be to enable the continual pre-training of these models, i.e. updating pre-trained models with new data instead of re-training them from scratch. However, the distribution shift induced by novel data typically results in degraded performance on past data. Taking a step towards efficient continual pre-training, in this work, we examine the effect of different warm-up strategies. Our hypothesis is that the learning rate must be re-increased to improve compute efficiency when training on a new dataset. We study the warmup phase of models pre-trained on the Pile (upstream data, 300B tokens) as we continue to pre-train on SlimPajama (downstream data, 297B tokens), following a linear warmup and cosine decay schedule. We conduct all experiments on the Pythia 410M language model architecture and evaluate performance through validation perplexity. We experiment with different pre-training checkpoints, various maximum learning rates, and various warmup lengths. Our results show that while rewarming models first increases the loss on upstream and downstream data, in the longer run it improves the downstream performance, outperforming models trained from scratch$\unicode{x2013}$even for a large downstream dataset.

One of the essential components of deep learning is the choice of the loss function and performance metrics used to train and evaluate models. This paper reviews the most prevalent loss functions and performance measurements in deep learning. We examine the benefits and limits of each technique and illustrate their application to various deep-learning problems. Our review aims to give a comprehensive picture of the different loss functions and performance indicators used in the most common deep learning tasks and help practitioners choose the best method for their specific task.

We applied physics-informed neural networks to solve the constitutive relations for nonlinear, path-dependent material behavior. As a result, the trained network not only satisfies all thermodynamic constraints but also instantly provides information about the current material state (i.e., free energy, stress, and the evolution of internal variables) under any given loading scenario without requiring initial data. One advantage of this work is that it bypasses the repetitive Newton iterations needed to solve nonlinear equations in complex material models. Additionally, strategies are provided to reduce the required order of derivative for obtaining the tangent operator. The trained model can be directly used in any finite element package (or other numerical methods) as a user-defined material model. However, challenges remain in the proper definition of collocation points and in integrating several non-equality constraints that become active or non-active simultaneously. We tested this methodology on rate-independent processes such as the classical von Mises plasticity model with a nonlinear hardening law, as well as local damage models for interface cracking behavior with a nonlinear softening law. In order to demonstrate the applicability of the methodology in handling complex path dependency in a three-dimensional (3D) scenario, we tested the approach using the equations governing a damage model for a three-dimensional interface model. Such models are frequently employed for intergranular fracture at grain boundaries. We have observed a perfect agreement between the results obtained through the proposed methodology and those obtained using the classical approach. Furthermore, the proposed approach requires significantly less effort in terms of implementation and computing time compared to the traditional methods.

Modern vision models typically rely on fine-tuning general-purpose models pre-trained on large, static datasets. These general-purpose models only capture the knowledge within their pre-training datasets, which are tiny, out-of-date snapshots of the Internet -- where billions of images are uploaded each day. We suggest an alternate approach: rather than hoping our static datasets transfer to our desired tasks after large-scale pre-training, we propose dynamically utilizing the Internet to quickly train a small-scale model that does extremely well on the task at hand. Our approach, called Internet Explorer, explores the web in a self-supervised manner to progressively find relevant examples that improve performance on a desired target dataset. It cycles between searching for images on the Internet with text queries, self-supervised training on downloaded images, determining which images were useful, and prioritizing what to search for next. We evaluate Internet Explorer across several datasets and show that it outperforms or matches CLIP oracle performance by using just a single GPU desktop to actively query the Internet for 30--40 hours. Results, visualizations, and videos at https://internet-explorer-ssl.github.io/

Accurate and real-time traffic state prediction is of great practical importance for urban traffic control and web mapping services. With the support of massive data, deep learning methods have shown their powerful capability in capturing the complex spatialtemporal patterns of traffic networks. However, existing approaches use pre-defined graphs and a simple set of spatial-temporal components, making it difficult to model multi-scale spatial-temporal dependencies. In this paper, we propose a novel dynamic graph convolution network with attention fusion to tackle this gap. The method first enhances the interaction of temporal feature dimensions, and then it combines a dynamic graph learner with GRU to jointly model synchronous spatial-temporal correlations. We also incorporate spatial-temporal attention modules to effectively capture longrange, multifaceted domain spatial-temporal patterns. We conduct extensive experiments in four real-world traffic datasets to demonstrate that our method surpasses state-of-the-art performance compared to 18 baseline methods.

Abstract--Unmanned aerial vehicles (UAVs) are frequently used for aerial mapping and general monitoring tasks. Recent progress in deep learning enabled automated semantic segmentation of imagery to facilitate the interpretation of large-scale complex environments. Commonly used supervised deep learning for segmentation relies on large amounts of pixel-wise labelled data, which is tedious and costly to annotate. The domain-specific visual appearance of aerial environments often prevents the usage of models pre-trained on publicly available datasets. To address this, we propose a novel general planning framework for UAVs to autonomously acquire informative training images for model re-training. Our framework combines the mapped acquisition function information into the UAV's planning objectives. In this way, the UAV adaptively acquires informative aerial images to be manually labelled for model re-training. Experimental results on real-world data and in a photorealistic simulation show that our framework maximises model performance and drastically reduces labelling efforts. Our map-based planners outperform state-of-the-art local planning. Our map-based planners replan a UAV's path (orange, bottom-left) to collect the most informative, e.g. Combined with advances in deep learning for semantic segmentation through fully convolutional improve the robot's vision capabilities in initially unknown neural networks (FCNs) [9, 10], deploying UAVs accelerates environments while minimising the total amount of humanlabelled automated scene understanding in large-scale and complex data. To this end, our approach exploits ideas from aerial environments [11]. Classical deep learning-based semantic AL research and incorporates them into a new informative segmentation models often used in this context are path planning (IPP) framework.

Graph convolutional networks (GCNs) have achieved great success in graph representation learning by extracting high-level features from nodes and their topology. Since GCNs generally follow a message-passing mechanism, each node aggregates information from its first-order neighbour to update its representation. As a result, the representations of nodes with edges between them should be positively correlated and thus can be considered positive samples. However, there are more non-neighbour nodes in the whole graph, which provide diverse and useful information for the representation update. Two non-adjacent nodes usually have different representations, which can be seen as negative samples. Besides the node representations, the structural information of the graph is also crucial for learning. In this paper, we used quality-diversity decomposition in determinant point processes (DPP) to obtain diverse negative samples. When defining a distribution on diverse subsets of all non-neighbouring nodes, we incorporate both graph structure information and node representations. Since the DPP sampling process requires matrix eigenvalue decomposition, we propose a new shortest-path-base method to improve computational efficiency. Finally, we incorporate the obtained negative samples into the graph convolution operation. The ideas are evaluated empirically in experiments on node classification tasks. These experiments show that the newly proposed methods not only improve the overall performance of standard representation learning but also significantly alleviate over-smoothing problems.

The issue of shortcut learning is widely known in NLP and has been an important research focus in recent years. Unintended correlations in the data enable models to easily solve tasks that were meant to exhibit advanced language understanding and reasoning capabilities. In this survey paper, we focus on the field of machine reading comprehension (MRC), an important task for showcasing high-level language understanding that also suffers from a range of shortcuts. We summarize the available techniques for measuring and mitigating shortcuts and conclude with suggestions for further progress in shortcut research. Importantly, we highlight two concerns for shortcut mitigation in MRC: (1) the lack of public challenge sets, a necessary component for effective and reusable evaluation, and (2) the lack of certain mitigation techniques that are prominent in other areas.