Artificial intelligence is not only changing the face of technology as we know it but also causing a seismic shift in the workforce and the talent and skills that the younger workforce, in particular, needs to own and exhibit in order to survive an ever-evolving technological future. AI in its current form has its origins in the field of education, but ironically, it is changing the very education system that gave birth to it. Computer science is no longer a unified field that encompasses a small body of scientific study. It has burgeoned into multiple disciplines that are entire fields of study in their own right. For example, at one point in time, neural networks was a highly specialized field that only the most dedicated computer science graduates would dare to enter.

In this issue of Maine Policy Review, authors provided insights into pressing concerns for Maine's policymakers, business leaders, educators, and citizens. They highlight research regarding critical challenges faced by Maine--and the nation--in industry, workforce, health, education, and politics. This "Maine Policy Minute" provides a synopsis of the authors and arguments contained in Volume 28, Number 1 of Maine Policy Review. In, "Local Politics from Away," transplanted Mainer and college student Matthew Bourque reflects on the strength and character of Maine's political traditions including independent thinking and politicians who put ideas over political party. Joseph W. McDonnell, a professor of public policy in management at the Muskie School at the University of Southern Maine, argues in "Maine's Workforce Challenges in an Age of Artificial Intelligence" that to accommodate changes wrought by increased automation and the move towards artificial intelligence, Maine needs to upgrade the skills of its workforce in what is becoming a rapidly changing economy.

When a robot acquires new information, ideally it would immediately be capable of using that information to understand its environment. While deep neural networks are now widely used by robots for inferring semantic information, conventional neural networks suffer from catastrophic forgetting when they are incrementally updated, with new knowledge overwriting established representations. While a variety of approaches have been developed that attempt to mitigate catastrophic forgetting in the incremental batch learning scenario, in which an agent learns a large collection of labeled samples at once, streaming learning has been much less studied in the robotics and deep learning communities. In streaming learning, an agent learns instances one-by-one and can be tested at any time. Here, we revisit streaming linear discriminant analysis, which has been widely used in the data mining research community. By combining streaming linear discriminant analysis with deep learning, we are able to outperform both incremental batch learning and streaming learning algorithms on both ImageNet-1K and CORe50.

Model extraction allows an adversary to steal a copy of a remotely deployed machine learning model given access to its predictions. Adversaries are motivated to mount such attacks for a variety of reasons, ranging from reducing their computational costs, to eliminating the need to collect expensive training data, to obtaining a copy of a model in order to find adversarial examples, perform membership inference, or model inversion attacks. In this paper, we taxonomize the space of model extraction attacks around two objectives: \emph{accuracy}, i.e., performing well on the underlying learning task, and \emph{fidelity}, i.e., matching the predictions of the remote victim classifier on any input. To extract a high-accuracy model, we develop a learning-based attack which exploits the victim to supervise the training of an extracted model. Through analytical and empirical arguments, we then explain the inherent limitations that prevent any learning-based strategy from extracting a truly high-fidelity model---i.e., extracting a functionally-equivalent model whose predictions are identical to those of the victim model on all possible inputs. Addressing these limitations, we expand on prior work to develop the first practical functionally-equivalent extraction attack for direct extraction (i.e., without training) of a model's weights. We perform experiments both on academic datasets and a state-of-the-art image classifier trained with 1 billion proprietary images. In addition to broadening the scope of model extraction research, our work demonstrates the practicality of model extraction attacks against production-grade systems.

Recently, consistency-based methods have achieved state-of-the-art results in semi-supervised learning (SSL). These methods always involve two roles, an explicit or implicit teacher model and a student model, and penalize predictions under different perturbations by a consistency constraint. However, the weights of these two roles are tightly coupled since the teacher is essentially an exponential moving average (EMA) of the student. In this work, we show that the coupled EMA teacher causes a performance bottleneck. To address this problem, we introduce Dual Student, which replaces the teacher with another student. We also define a novel concept, stable sample, following which a stabilization constraint is designed for our structure to be trainable. Further, we discuss two variants of our method, which produce even higher performance. Extensive experiments show that our method improves the classification performance significantly on several main SSL benchmarks. Specifically, it reduces the error rate of the 13-layer CNN from 16.84% to 12.39% on CIFAR-10 with 1k labels and from 34.10% to 31.56% on CIFAR-100 with 10k labels. In addition, our method also achieves a clear improvement in domain adaptation.

Neural networks enjoy widespread use, but many aspects of their training, representation, and operation are poorly understood. In particular, our view into the training process is limited, with a single scalar loss being the most common viewport into this high-dimensional, dynamic process. We propose a new window into training called Loss Change Allocation (LCA), in which credit for changes to the network loss is conservatively partitioned to the parameters. This measurement is accomplished by decomposing the components of an approximate path integral along the training trajectory using a Runge-Kutta integrator. This rich view shows which parameters are responsible for decreasing or increasing the loss during training, or which parameters "help" or "hurt" the network's learning, respectively. LCA may be summed over training iterations and/or over neurons, channels, or layers for increasingly coarse views. This new measurement device produces several insights into training. (1) We find that barely over 50% of parameters help during any given iteration. (2) Some entire layers hurt overall, moving on average against the training gradient, a phenomenon we hypothesize may be due to phase lag in an oscillatory training process. (3) Finally, increments in learning proceed in a synchronized manner across layers, often peaking on identical iterations.

Research on graph representation learning has received a lot of attention in recent years since many data in real-world applications come in form of graphs. High-dimensional graph data are often in irregular form, which makes them more difficult to analyze than image/video/audio data defined on regular lattices. Various graph embedding techniques have been developed to convert the raw graph data into a low-dimensional vector representation while preserving the intrinsic graph properties. In this review, we first explain the graph embedding task and its challenges. Next, we review a wide range of graph embedding techniques with insights. Then, we evaluate several state-of-the-art methods against small and large datasets and compare their performance. Finally, potential applications and future directions are presented.

Classifiers that achieve demographic balance by explicitly using protected attributes such as race or gender are often politically or culturally controversial due to their lack of individual fairness, i.e. individuals with similar qualifications will receive different outcomes. Individually and group fair decision criteria can produce counter-intuitive results, e.g. that the optimal constrained boundary may reject intuitively better candidates due to demographic imbalance in similar candidates. Both approaches can be seen as introducing individual resentment, where some individuals would have received a better outcome if they either belonged to a different demographic class and had the same qualifications, or if they remained in the same class but had objectively worse qualifications (e.g. lower test scores). We show that both forms of resentment can be avoided by using monotonically constrained machine learning models to create individually fair, demographically balanced classifiers.

Agents trained with deep reinforcement learning algorithms are capable of performing highly complex tasks including locomotion in continuous environments. In order to attain a human-level performance, the next step of research should be to investigate the ability to transfer the learning acquired in one task to a different set of tasks. Concerns on generalization and overfitting in deep reinforcement learning are not usually addressed in current transfer learning research. This issue results in underperforming benchmarks and inaccurate algorithm comparisons due to rudimentary assessments. In this study, we primarily propose regularization techniques in deep reinforcement learning for continuous control through the application of sample elimination and early stopping. First, the importance of the inclusion of training iteration to the hyperparameters in deep transfer learning problems will be emphasized. Because source task performance is not indicative of the generalization capacity of the algorithm, we start by proposing various transfer learning evaluation methods that acknowledge the training iteration as a hyperparameter. In line with this, we introduce an additional step of resorting to earlier snapshots of policy parameters depending on the target task due to overfitting to the source task. Then, in order to generate robust policies,we discard the samples that lead to overfitting via strict clipping. Furthermore, we increase the generalization capacity in widely used transfer learning benchmarks by using entropy bonus, different critic methods and curriculum learning in an adversarial setup. Finally, we evaluate the robustness of these techniques and algorithms on simulated robots in target environments where the morphology of the robot, gravity and tangential friction of the environment are altered from the source environment.

Reasoning over plots by question answering (QA) is a challenging machine learning task at the intersection of vision, language processing, and reasoning. Existing synthetic datasets (FigureQA, DVQA) do not model variability in data labels, real-valued data, or complex reasoning questions. Consequently, proposed models for these datasets do not fully address the challenge of reasoning over plots. We propose PlotQA with 8.1 million question-answer pairs over 220,000 plots with data from real-world sources and questions based on crowd-sourced question templates. 26% of the questions in PlotQA have answers that are not in a fixed vocabulary, requiring reasoning capabilities. Analysis of existing models on PlotQA reveals that a hybrid model is required: Specific questions are answered better by choosing the answer from a fixed vocabulary or by extracting it from a predicted bounding box in the plot, while other questions are answered with a table question-answering engine which is fed with a structured table extracted by visual element detection. For the latter, we propose the VOES pipeline and combine it with SAN-VQA to form a hybrid model SAN-VOES. On the DVQA dataset, SAN-VOES model has an accuracy of 58%, significantly improving on highest reported accuracy of 46%. On the PlotQA dataset, SAN-VOES has an accuracy of 54%, which is the highest amongst all the models we trained. Analysis of each module in the VOES pipeline reveals that further improvement in accuracy requires more accurate visual element detection.