We introduce a large language model (LLM) based approach to answer complex questions requiring multi-hop numerical reasoning over financial reports. While LLMs have exhibited remarkable performance on various natural language and reasoning tasks, complex reasoning problems often rely on few-shot prompts that require carefully crafted examples. In contrast, our approach uses novel zero-shot prompts that guide the LLM to encode the required reasoning into a Python program or a domain specific language. The generated program is then executed by a program interpreter, thus mitigating the limitations of LLM in performing accurate arithmetic calculations. We evaluate the proposed approach on three financial datasets using some of the recently developed generative pretrained transformer (GPT) models and perform comparisons with various zero-shot baselines. The experimental results demonstrate that our approach significantly improves the accuracy for all the LLMs over their respective baselines. We provide a detailed analysis of the results, generating insights to support our findings. The success of our approach demonstrates the enormous potential to extract complex domain specific numerical reasoning by designing zero-shot prompts to effectively exploit the knowledge embedded in LLMs.

Data-driven models created by machine learning gain in importance in all fields of design and engineering. They have high potential to assist decision-makers in creating novel artefacts with better performance and sustainability. However, limited generalization and the black-box nature of these models lead to limited explainability and reusability. To overcome this situation, we propose a component-based approach to create partial component models by machine learning (ML). This component-based approach aligns deep learning with systems engineering (SE). For the domain of energy efficient building design, we first demonstrate better generalization of the component-based method by analyzing prediction accuracy outside the training data. We observe a much higher accuracy (R2 = 0.94) compared to conventional monolithic methods (R2 = 0.71). Second, we illustrate explainability by exemplary demonstrating how sensitivity information from SE and rules from low-depth decision trees serve engineering. Third, we evaluate explainability by qualitative and quantitative methods demonstrating the matching of preliminary knowledge and data-driven derived strategies and show correctness of activations at component interfaces compared to white-box simulation results (envelope components: R2 = 0.92..0.99; zones: R2 = 0.78..0.93). The key for component-based explainability is that activations at interfaces between the components are interpretable engineering quantities. The large range of possible configurations in composing components allows the examination of novel unseen design cases with understandable data-driven models. The matching of parameter ranges of components by similar probability distribution produces reusable, well-generalizing, and trustworthy models. The approach adapts the model structure to engineering methods of systems engineering and to domain knowledge.

With the rapid growth of Artificial Intelligence (AI) applications supported by deep learning (DL), the energy efficiency of these applications has an increasingly large impact on sustainability. We introduce Smaragdine, a new energy accounting system for tensor-based DL programs implemented with TensorFlow. At the heart of Smaragdine is a novel white-box methodology of energy accounting: Smaragdine is aware of the internal structure of the DL program, which we call tensor-aware energy accounting. With Smaragdine, the energy consumption of a DL program can be broken down into units aligned with its logical hierarchical decomposition structure. We apply Smaragdine for understanding the energy behavior of BERT, one of the most widely used language models. Layer-by-layer and tensor-by-tensor, Smaragdine is capable of identifying the highest energy/power-consuming components of BERT. Furthermore, we conduct two case studies on how Smaragdine supports downstream toolchain building, one on the comparative energy impact of hyperparameter tuning of BERT, the other on the energy behavior evolution when BERT evolves to its next generation, ALBERT.

Self-supervised learning for time-series data holds potential similar to that recently unleashed in Natural Language Processing and Computer Vision. While most existing works in this area focus on contrastive learning, we propose a conceptually simple yet powerful non-contrastive approach, based on the data2vec self-distillation framework. The core of our method is a student-teacher scheme that predicts the latent representation of an input time series from masked views of the same time series. This strategy avoids strong modality-specific assumptions and biases typically introduced by the design of contrastive sample pairs. We demonstrate the competitiveness of our approach for classification and forecasting as downstream tasks, comparing with state-of-the-art self-supervised learning methods on the UCR and UEA archives as well as the ETT and Electricity datasets.

We present a review that unifies decision-support methods for exploring the solutions produced by multi-objective optimization (MOO) algorithms. As MOO is applied to solve diverse problems, approaches for analyzing the trade-offs offered by MOO algorithms are scattered across fields. We provide an overview of the advances on this topic, including methods for visualization, mining the solution set, and uncertainty exploration as well as emerging research directions, including interactivity, explainability, and ethics. We synthesize these methods drawing from different fields of research to build a unified approach, independent of the application. Our goals are to reduce the entry barrier for researchers and practitioners on using MOO algorithms and to provide novel research directions.

Time series is a special type of sequence data, a sequence of real-valued random variables collected at even intervals of time. The real-world multivariate time series comes with noises and contains complicated local and global temporal dynamics, making it difficult to forecast the future time series given the historical observations. This work proposes a simple and effective framework, coined as TimeSQL, which leverages multi-scale patching and smooth quadratic loss (SQL) to tackle the above challenges. The multi-scale patching transforms the time series into two-dimensional patches with different length scales, facilitating the perception of both locality and long-term correlations in time series. SQL is derived from the rational quadratic kernel and can dynamically adjust the gradients to avoid overfitting to the noises and outliers. Theoretical analysis demonstrates that, under mild conditions, the effect of the noises on the model with SQL is always smaller than that with MSE. Based on the two modules, TimeSQL achieves new state-of-the-art performance on the eight real-world benchmark datasets. Further ablation studies indicate that the key modules in TimeSQL could also enhance the results of other models for multivariate time series forecasting, standing as plug-and-play techniques.

Deep learning has shown promising performance in submeter-level mapping tasks; however, the annotation cost of submeter-level imagery remains a challenge, especially when applied on a large scale. In this paper, we present the first submeter-level land cover mapping of Japan with eight classes, at a relatively low annotation cost. We introduce a human-in-the-loop deep learning framework leveraging OpenEarthMap, a recently introduced benchmark dataset for global submeter-level land cover mapping, with a U-Net model that achieves national-scale mapping with a small amount of additional labeled data. By adding a small amount of labeled data of areas or regions where a U-Net model trained on OpenEarthMap clearly failed and retraining the model, an overall accuracy of 80\% was achieved, which is a nearly 16 percentage point improvement after retraining. Using aerial imagery provided by the Geospatial Information Authority of Japan, we create land cover classification maps of eight classes for the entire country of Japan. Our framework, with its low annotation cost and high-accuracy mapping results, demonstrates the potential to contribute to the automatic updating of national-scale land cover mapping using submeter-level optical remote sensing data. The mapping results will be made publicly available.

Numerous web applications rely on solving combinatorial optimization problems, such as energy cost-aware scheduling, budget allocation on web advertising, and graph matching on social networks. However, many optimization problems involve unknown coefficients, and improper predictions of these factors may lead to inferior decisions which may cause energy wastage, inefficient resource allocation, inappropriate matching in social networks, etc. Such a research topic is referred to as "Predict-Then-Optimize (PTO)" which considers the performance of prediction and decision-making in a unified system. A noteworthy recent development is the end-to-end methods by directly optimizing the ultimate decision quality which claims to yield better results in contrast to the traditional two-stage approach. However, the evaluation benchmarks in this field are fragmented and the effectiveness of various models in different scenarios remains unclear, hindering the comprehensive assessment and fast deployment of these methods. To address these issues, we provide a comprehensive categorization of current approaches and integrate existing experimental scenarios to establish a unified benchmark, elucidating the circumstances under which end-to-end training yields improvements, as well as the contexts in which it performs ineffectively. We also introduce a new dataset for the industrial combinatorial advertising problem for inclusive finance to open-source. We hope the rethinking and benchmarking of PTO could facilitate more convenient evaluation and deployment, and inspire further improvements both in the academy and industry within this field.

Ensuring the safety of reinforcement learning (RL) algorithms is crucial to unlock their potential for many real-world tasks. However, vanilla RL and most safe RL approaches do not guarantee safety. In recent years, several methods have been proposed to provide hard safety guarantees for RL, which is essential for applications where unsafe actions could have disastrous consequences. Nevertheless, there is no comprehensive comparison of these provably safe RL methods. Therefore, we introduce a categorization of existing provably safe RL methods, present the conceptual foundations for both continuous and discrete action spaces, and empirically benchmark existing methods. We categorize the methods based on how they adapt the action: action replacement, action projection, and action masking. Our experiments on an inverted pendulum and a quadrotor stabilization task indicate that action replacement is the best-performing approach for these applications despite its comparatively simple realization. Furthermore, adding a reward penalty, every time the safety verification is engaged, improved training performance in our experiments. Finally, we provide practical guidance on selecting provably safe RL approaches depending on the safety specification, RL algorithm, and type of action space.

Soft robotics is a swiftly evolving field. Pneumatic actuators are suitable for driving soft robots because of their superior performance. However, their control is challenging due to the hysteresis characteristics. In response to this challenge, we propose an adaptive control method to compensate for the hysteresis of soft actuators. Employing a novel dual pneumatic artificial muscle (PAM) bending actuator, the innovative control approach abates hysteresis effects by dynamically modulating gains within a traditional PID controller corresponding to the predicted variation of the reference trajectory. Through experimental evaluation, we found that the proposed control method outperforms its conventional counterparts regarding tracking accuracy and response speed. Our work reveals a new direction for advancing model-free control in soft actuators.