As mathematical, statistical, and machine learning algorithms leverage increasingly powerful computational hardware to perform elaborate tasks, simulation has grown to play a key role in fields such as materials science, operations research, industrial engineering, aerodynamics, pharmaceuticals, image processing, and many others. In particular, a key use of these simulations is to serve as a surrogate for the eventual implementation and/or manufacturing during the design optimization; running a computational simulation is likely much cheaper than actually conducting a physical experiment or fabrication (Forrester et al. 2008; Negoescu et al. 2011; Molesky et al. 2018; Haghanifar et al. 2020). Computational simulations can, however, easily run for hours or days, making simulation itself an often costly proposition. The high cost of a single simulation is compounded by the frequent need to simulate many different systems to search for a set of desirable outcomes. This is the motivating force behind simulation optimization, which seeks to identify suitable system parameters to achieve a satisfactory system or effective simulation in a sample-efficient fashion, i.e., with as few simulations conducted as possible. In practical situations, simulations almost always have multiple competing objectives which define success, and thus it is important for users to understand trade-offs between these competing objectives in order to make an informed design decision. Multiobjective optimization tackles this problem by identifying the Pareto frontier, which is the manifold in objective space such that improving one objective cannot occur without harming another. Unfortunately, using the Pareto frontier as the measurement of success may be limiting in engineering and design applications.

The Generative Adversarial Network (GAN) was recently introduced in the literature as a novel machine learning method for training generative models. It has many applications in statistics such as nonparametric clustering and nonparametric conditional independence tests. However, training the GAN is notoriously difficult due to the issue of mode collapse, which refers to the lack of diversity among generated data. In this paper, we identify the reasons why the GAN suffers from this issue, and to address it, we propose a new formulation for the GAN based on randomized decision rules. In the new formulation, the discriminator converges to a fixed point while the generator converges to a distribution at the Nash equilibrium. We propose to train the GAN by an empirical Bayes-like method by treating the discriminator as a hyper-parameter of the posterior distribution of the generator. Specifically, we simulate generators from its posterior distribution conditioned on the discriminator using a stochastic gradient Markov chain Monte Carlo (MCMC) algorithm, and update the discriminator using stochastic gradient descent along with simulations of the generators. We establish convergence of the proposed method to the Nash equilibrium. Apart from image generation, we apply the proposed method to nonparametric clustering and nonparametric conditional independence tests. A portion of the numerical results is presented in the supplementary material.

The capabilities and adoption of deep neural networks (DNNs) grow at an exhilarating pace: Vision models accurately classify human actions in videos and identify cancerous tissue in medical scans as precisely than human experts; large language models answer wide-ranging questions, generate code, and write prose, becoming the topic of everyday dinner-table conversations. Even though their uses are exhilarating, the continually increasing model sizes and computational complexities have a dark side. The economic cost and negative environmental externalities of training and serving models is in evident disharmony with financial viability and climate action goals. Instead of pursuing yet another increase in predictive performance, this dissertation is dedicated to the improvement of neural network efficiency. Specifically, a core contribution addresses the efficiency aspects during online inference. Here, the concept of Continual Inference Networks (CINs) is proposed and explored across four publications. CINs extend prior state-of-the-art methods developed for offline processing of spatio-temporal data and reuse their pre-trained weights, improving their online processing efficiency by an order of magnitude. These advances are attained through a bottom-up computational reorganization and judicious architectural modifications. The benefit to online inference is demonstrated by reformulating several widely used network architectures into CINs, including 3D CNNs, ST-GCNs, and Transformer Encoders. An orthogonal contribution tackles the concurrent adaptation and computational acceleration of a large source model into multiple lightweight derived models. Drawing on fusible adapter networks and structured pruning, Structured Pruning Adapters achieve superior predictive accuracy under aggressive pruning using significantly fewer learned weights compared to fine-tuning with pruning.

Abstract Meaning Representation (AMR) is a Semantic Parsing formalism that aims at providing a semantic graph abstraction representing a given text. Current approaches are based on autoregressive language models such as BART or T5, fine-tuned through Teacher Forcing to obtain a linearized version of the AMR graph from a sentence. In this paper, we present LeakDistill, a model and method that explores a modification to the Transformer architecture, using structural adapters to explicitly incorporate graph information into the learned representations and improve AMR parsing performance. Our experiments show how, by employing word-to-node alignment to embed graph structural information into the encoder at training time, we can obtain state-of-the-art AMR parsing through self-knowledge distillation, even without the use of additional data. We release the code at \url{http://www.github.com/sapienzanlp/LeakDistill}.

The concepts of overfitting and generalization are vital for evaluating machine learning models. In this work, we show that the popular Recall@K metric depends on the number of classes in the dataset, which limits its ability to estimate generalization. To fix this issue, we propose a new metric, which measures retrieval performance, and, unlike Recall@K, estimates generalization. We apply the proposed metric to popular image retrieval methods and provide new insights about deep metric learning generalization.

The utilization of teleoperation is a crucial aspect of the construction industry, as it enables operators to control machines safely from a distance. However, remote operation of these machines at a joint level using individual joysticks necessitates extensive training for operators to achieve proficiency due to their multiple degrees of freedom. Additionally, verifying the machine resulting motion is only possible after execution, making optimal control challenging. In addressing this issue, this study proposes a reinforcement learning-based approach to optimize task performance. The control policy acquired through learning is used to provide instructions on efficiently controlling and coordinating multiple joints. To evaluate the effectiveness of the proposed framework, a user study is conducted with a Brokk 170 construction machine by assessing its performance in a typical construction task involving inserting a chisel into a borehole. The effectiveness of the proposed framework is evaluated by comparing the performance of participants in the presence and absence of virtual fixtures. This study results demonstrate the proposed framework potential in enhancing the teleoperation process in the construction industry.

To fully exploit the benefits of the fog environment, efficient management of data locality is crucial. Blind or reactive data replication falls short in harnessing the potential of fog computing, necessitating more advanced techniques for predicting where and when clients will connect. While spatial prediction has received considerable attention, temporal prediction remains understudied. Our paper addresses this gap by examining the advantages of incorporating temporal prediction into existing spatial prediction models. We also provide a comprehensive analysis of spatio-temporal prediction models, such as Deep Neural Networks and Markov models, in the context of predictive replication. We propose a novel model using Holt-Winter's Exponential Smoothing for temporal prediction, leveraging sequential and periodical user movement patterns. In a fog network simulation with real user trajectories our model achieves a 15% reduction in excess data with a marginal 1% decrease in data availability.

Many robotic surgical systems have been developed with micro-sized biopsy forceps for tissue manipulation. However, these systems often lack force sensing at the tool side. This paper presents a vision-based force sensing method for micro-sized biopsy forceps. A miniature sensing module adaptive to common biopsy forceps is proposed, consisting of a flexure, a camera, and a customised target. The deformation of the flexure is obtained by the camera estimating the pose variation of the top-mounted target. Then, the external force applied to the sensing module is calculated using the flexure's displacement and stiffness matrix. Integrating the sensing module into the biopsy forceps, in conjunction with a single-axial force sensor at the proximal end, we equip the forceps with haptic sensing capabilities. Mathematical equations are derived to estimate the multi-modal force sensing of the haptics-enabled forceps, including pushing/pulling forces (Mode-I) and grasping forces (Mode-II). A series of experiments on phantoms and ex vivo tissues are conducted to verify the feasibility of the proposed design and method. Results indicate that the haptics-enabled forceps can achieve multi-modal force estimation effectively and potentially realize autonomous robotic tissue grasping procedures with controlled forces.

After just a few hundred training updates, a standard probabilistic model for language generation has likely not yet learnt many semantic or syntactic rules of natural language, making it difficult to estimate the probability distribution over next tokens. Yet around this point, these models have identified a simple, loss-minimising behaviour: to output the unigram distribution of the target training corpus. The use of such a heuristic raises the question: Can we initialise our models with this behaviour and save precious compute resources and model capacity? Here we show that we can effectively endow standard neural language generation models with a separate module that reflects unigram frequency statistics as prior knowledge, simply by initialising the bias term in a model's final linear layer with the log-unigram distribution. We use neural machine translation as a test bed for this simple technique and observe that it: (i) improves learning efficiency; (ii) achieves better overall performance; and perhaps most importantly (iii) appears to disentangle strong frequency effects by encouraging the model to specialise in non-frequency-related aspects of language.

Due to the usefulness in data enrichment for data analysis tasks, joinable table discovery has become an important operation in data lake management. Existing approaches target equi-joins, the most common way of combining tables for creating a unified view, or semantic joins, which tolerate misspellings and different formats to deliver more join results. They are either exact solutions whose running time is linear in the sizes of query column and target table repository or approximate solutions lacking precision. In this paper, we propose Deepjoin, a deep learning model for accurate and efficient joinable table discovery. Our solution is an embedding-based retrieval, which employs a pre-trained language model (PLM) and is designed as one framework serving both equi- and semantic joins. We propose a set of contextualization options to transform column contents to a text sequence. The PLM reads the sequence and is fine-tuned to embed columns to vectors such that columns are expected to be joinable if they are close to each other in the vector space. Since the output of the PLM is fixed in length, the subsequent search procedure becomes independent of the column size. With a state-of-the-art approximate nearest neighbor search algorithm, the search time is logarithmic in the repository size. To train the model, we devise the techniques for preparing training data as well as data augmentation. The experiments on real datasets demonstrate that by training on a small subset of a corpus, Deepjoin generalizes to large datasets and its precision consistently outperforms other approximate solutions'. Deepjoin is even more accurate than an exact solution to semantic joins when evaluated with labels from experts. Moreover, when equipped with a GPU, Deepjoin is up to two orders of magnitude faster than existing solutions.