Dealing with missing data is an important problem in statistical analysis that is often addressed with imputation procedures. The performance and validity of such methods are of great importance for their application in empirical studies. While the prevailing method of Multiple Imputation by Chained Equations (MICE) with Predictive Mean Matching (PMM) is considered standard in the social science literature, the increase in complex datasets may require more advanced approaches based on machine learning. In particular, tree-based imputation methods have emerged as very competitive approaches. However, the performance and validity are not completely understood, particularly compared to the standard MICE PMM. This is especially true for inference in linear models. In this study, we investigate the impact of various imputation methods on coefficient estimation, Type I error, and power, to gain insights that can help empirical researchers deal with missingness more effectively. We explore MICE PMM alongside different tree-based methods, such as MICE with Random Forest (RF), Chained Random Forests with and without PMM (missRanger), and Extreme Gradient Boosting (MIXGBoost), conducting a realistic simulation study using the German National Educational Panel Study (NEPS) as the original data source. Our results reveal that Random Forest-based imputations, especially MICE RF and missRanger with PMM, consistently perform better in most scenarios. Standard MICE PMM shows partially increased bias and overly conservative test decisions, particularly with non-true zero coefficients. Our results thus underscore the potential advantages of tree-based imputation methods, albeit with a caveat that all methods perform worse with an increased missingness, particularly missRanger.

The Curriculum Recommendations paradigm is dedicated to fostering learning equality within the ever-evolving realms of educational technology and curriculum development. In acknowledging the inherent obstacles posed by existing methodologies, such as content conflicts and disruptions from language translation, this paradigm aims to confront and overcome these challenges. Notably, it addresses content conflicts and disruptions introduced by language translation, hindrances that can impede the creation of an all-encompassing and personalized learning experience. The paradigm's objective is to cultivate an educational environment that not only embraces diversity but also customizes learning experiences to suit the distinct needs of each learner. To overcome these challenges, our approach builds upon notable contributions in curriculum development and personalized learning, introducing three key innovations. These include the integration of Transformer Base Model to enhance computational efficiency, the implementation of InfoNCE Loss for accurate content-topic matching, and the adoption of a language switching strategy to alleviate translation-related ambiguities. Together, these innovations aim to collectively tackle inherent challenges and contribute to forging a more equitable and effective learning journey for a diverse range of learners. Competitive cross-validation scores underscore the efficacy of sentence-transformers/LaBSE, achieving 0.66314, showcasing our methodology's effectiveness in diverse linguistic nuances for content alignment prediction. Index Terms-Curriculum Recommendation, Transformer model with InfoNCE Loss, Language Switching.

Relationships are essential to our happiness and wellbeing. The dissolution of a relationship, the final stage of relationship's lifecycle and one of the most stressful events in an individual's life, can have profound and long-lasting impacts on people. With the breakup process increasingly facilitated by computer-mediated communication (CMC), and the likely future influence of AI-mediated communication (AIMC) tools, we conducted a semi-structured interview study with 21 participants. We aim to understand: 1) the current role of technology in the breakup process, 2) the needs and support individuals have during the process, and 3) how AI might address these needs. Our research shows that people have distinct needs at various stages of ending a relationship. Presently, technology is used for information gathering and community support, acting as a catalyst for breakups, enabling ghosting and blocking, and facilitating communication. Participants anticipate that AI could aid in sense-making of their relationship leading up to the breakup, act as a mediator, assist in crafting appropriate wording, tones, and language during breakup conversations, and support companionship, reflection, recovery, and growth after a breakup. Our findings also demonstrate an overlap between the breakup process and the Transtheoretical Model (TTM) of behavior change. Through the lens of TTM, we explore the potential support and affordances AI could offer in breakups, including its benefits and the necessary precautions regarding AI's role in this sensitive process.

Categorical variables often appear in datasets for classification and regression tasks, and they need to be encoded into numerical values before training. Since many encoders have been developed and can significantly impact performance, choosing the appropriate encoder for a task becomes a time-consuming yet important practical issue. This study broadly classifies machine learning models into three categories: 1) ATI models that implicitly perform affine transformations on inputs, such as multi-layer perceptron neural network; 2) Tree-based models that are based on decision trees, such as random forest; and 3) the rest, such as kNN. Theoretically, we prove that the one-hot encoder is the best choice for ATI models in the sense that it can mimic any other encoders by learning suitable weights from the data. We also explain why the target encoder and its variants are the most suitable encoders for tree-based models. This study conducted comprehensive computational experiments to evaluate 14 encoders, including one-hot and target encoders, along with eight common machine-learning models on 28 datasets. The computational results agree with our theoretical analysis. The findings in this study shed light on how to select the suitable encoder for data scientists in fields such as fraud detection, disease diagnosis, etc.

Hierarchical federated learning (HFL) enables distributed training of models across multiple devices with the help of several edge servers and a cloud edge server in a privacy-preserving manner. In this paper, we consider HFL with highly mobile devices, mainly targeting at vehicular networks. Through convergence analysis, we show that mobility influences the convergence speed by both fusing the edge data and shuffling the edge models. While mobility is usually considered as a challenge from the perspective of communication, we prove that it increases the convergence speed of HFL with edge-level heterogeneous data, since more diverse data can be incorporated. Furthermore, we demonstrate that a higher speed leads to faster convergence, since it accelerates the fusion of data. Simulation results show that mobility increases the model accuracy of HFL by up to 15.1% when training a convolutional neural network on the CIFAR-10 dataset.

This paper introduces ClimateGPT, a model family of domain-specific large language models that synthesize interdisciplinary research on climate change. We trained two 7B models from scratch on a science-oriented dataset of 300B tokens. For the first model, the 4.2B domain-specific tokens were included during pre-training and the second was adapted to the climate domain after pre-training. Additionally, ClimateGPT-7B, 13B and 70B are continuously pre-trained from Llama~2 on a domain-specific dataset of 4.2B tokens. Each model is instruction fine-tuned on a high-quality and human-generated domain-specific dataset that has been created in close cooperation with climate scientists. To reduce the number of hallucinations, we optimize the model for retrieval augmentation and propose a hierarchical retrieval strategy. To increase the accessibility of our model to non-English speakers, we propose to make use of cascaded machine translation and show that this approach can perform comparably to natively multilingual models while being easier to scale to a large number of languages. Further, to address the intrinsic interdisciplinary aspect of climate change we consider different research perspectives. Therefore, the model can produce in-depth answers focusing on different perspectives in addition to an overall answer. We propose a suite of automatic climate-specific benchmarks to evaluate LLMs. On these benchmarks, ClimateGPT-7B performs on par with the ten times larger Llama-2-70B Chat model while not degrading results on general domain benchmarks. Our human evaluation confirms the trends we saw in our benchmarks. All models were trained and evaluated using renewable energy and are released publicly.

Hyper-parameter optimization is the black art of tuning a learner's control parameters. In software analytics, a repeated result is that such tuning can result in dramatic performance improvements. Despite this, hyper-parameter optimization is often applied rarely or poorly in software analytics--perhaps due to the CPU cost of exploring all those parameter options can be prohibitive. We theorize that learners generalize better when the loss landscape is ``smooth''. This theory is useful since the influence on ``smoothness'' of different hyper-parameter choices can be tested very quickly (e.g. for a deep learner, after just one epoch). To test this theory, this paper implements and tests SMOOTHIE, a novel hyper-parameter optimizer that guides its optimizations via considerations of ``smothness''. The experiments of this paper test SMOOTHIE on numerous SE tasks including (a) GitHub issue lifetime prediction; (b) detecting false alarms in static code warnings; (c) defect prediction, and (d) a set of standard ML datasets. In all these experiments, SMOOTHIE out-performed state-of-the-art optimizers. Better yet, SMOOTHIE ran 300% faster than the prior state-of-the art. We hence conclude that this theory (that hyper-parameter optimization is best viewed as a ``smoothing'' function for the decision landscape), is both theoretically interesting and practically very useful. To support open science and other researchers working in this area, all our scripts and datasets are available on-line at https://github.com/yrahul3910/smoothness-hpo/.

Bilevel optimization is an important formulation for many machine learning problems, such as meta-learning and hyperparameter optimization. Current bilevel optimization algorithms assume that the gradient of the upper-level function is Lipschitz (i.e., the upper-level function has a bounded smoothness parameter). However, recent studies reveal that certain neural networks such as recurrent neural networks (RNNs) and long-short-term memory networks (LSTMs) exhibit potential unbounded smoothness, rendering conventional bilevel optimization algorithms unsuitable for these neural networks. In this paper, we design a new bilevel optimization algorithm, namely BO-REP, to address this challenge. This algorithm updates the upper-level variable using normalized momentum and incorporates two novel techniques for updating the lower-level variable: initialization refinement and periodic updates. Specifically, once the upper-level variable is initialized, a subroutine is invoked to obtain a refined estimate of the corresponding optimal lower-level variable, and the lower-level variable is updated only after every specific period instead of each iteration. Notably, this result matches the state-of-the-art complexity results under the bounded smoothness setting and without mean-squared smoothness of the stochastic gradient, up to logarithmic factors. Our proof relies on novel technical lemmas for the periodically updated lower-level variable, which are of independent interest. Our experiments on hyperrepresentation learning, hyperparameter optimization, and data hyper-cleaning for text classification tasks demonstrate the effectiveness of our proposed algorithm. Bilevel optimization refers to an optimization problem where one problem is nested within another (Bracken & McGill, 1973; Dempe, 2002). One important application under this setting is hyper-representation learning with deep neural networks (Franceschi et al., 2018), where x denotes the shared representation Table 1: Comparison of oracle complexity of stochastic bilevel algorithms for finding an ϵ-stationary point as defined in Definition 1. The oracle stands for stochastic gradient and stochastic Hessian vector product.

We study the benefit of sharing representations among tasks to enable the effective use of deep neural networks in Multi-Task Reinforcement Learning. We leverage the assumption that learning from different tasks, sharing common properties, is helpful to generalize the knowledge of them resulting in a more effective feature extraction compared to learning a single task. Intuitively, the resulting set of features offers performance benefits when used by Reinforcement Learning algorithms. We prove this by providing theoretical guarantees that highlight the conditions for which is convenient to share representations among tasks, extending the wellknown finite-time bounds of Approximate Value-Iteration to the multi-task setting. In addition, we complement our analysis by proposing multi-task extensions of three Reinforcement Learning algorithms that we empirically evaluate on widely used Reinforcement Learning benchmarks showing significant improvements over the single-task counterparts in terms of sample efficiency and performance. Multi-Task Learning (MTL) ambitiously aims to learn multiple tasks jointly instead of learning them separately, leveraging the assumption that the considered tasks have common properties which can be exploited by Machine Learning (ML) models to generalize the learning of each of them. For instance, the features extracted in the hidden layers of a neural network trained on multiple tasks have the advantage of being a general representation of structures common to each other.

3D vision-language grounding, which focuses on aligning language with the 3D physical environment, stands as a cornerstone in the development of embodied agents. In comparison to recent advancements in the 2D domain, grounding language in 3D scenes faces several significant challenges: (i) the inherent complexity of 3D scenes due to the diverse object configurations, their rich attributes, and intricate relationships; (ii) the scarcity of paired 3D vision-language data to support grounded learning; and (iii) the absence of a unified learning framework to distill knowledge from grounded 3D data. In this work, we aim to address these three major challenges in 3D vision-language by examining the potential of systematically upscaling 3D vision-language learning in indoor environments. We introduce the first million-scale 3D vision-language dataset, SceneVerse, encompassing about 68K 3D indoor scenes and comprising 2.5M vision-language pairs derived from both human annotations and our scalable scene-graph-based generation approach. We demonstrate that this scaling allows for a unified pre-training framework, Grounded Pre-training for Scenes (GPS), for 3D vision-language learning. Through extensive experiments, we showcase the effectiveness of GPS by achieving state-of-the-art performance on all existing 3D visual grounding benchmarks. The vast potential of SceneVerse and GPS is unveiled through zero-shot transfer experiments in the challenging 3D vision-language tasks. Project website: https://scene-verse.github.io .