Weight decay is a popular regularization technique for training of deep neural networks. Modern deep learning libraries mainly use $L_{2}$ regularization as the default implementation of weight decay. \citet{loshchilov2018decoupled} demonstrated that $L_{2}$ regularization is not identical to weight decay for adaptive gradient methods, such as Adaptive Momentum Estimation (Adam), and proposed Adam with Decoupled Weight Decay (AdamW). However, we found that the popular implementations of weight decay, including $L_{2}$ regularization and decoupled weight decay, in modern deep learning libraries usually damage performance. First, the $L_{2}$ regularization is unstable weight decay for all optimizers that use Momentum, such as stochastic gradient descent (SGD). Second, decoupled weight decay is highly unstable for all adaptive gradient methods. We further propose the Stable Weight Decay (SWD) method to fix the unstable weight decay problem from a dynamical perspective. The proposed SWD method makes significant improvements over $L_{2}$ regularization and decoupled weight decay in our experiments. Simply fixing weight decay in Adam by SWD, with no extra hyperparameter, can usually outperform complex Adam variants, which have more hyperparameters.

There exists an apparent negative correlation between performance and interpretability of deep learning models. In an effort to reduce this negative correlation, we propose Born Identity Network (BIN), which is a post-hoc approach for producing multi-way counterfactual maps. A counterfactual map transforms an input sample to be classified as a target label, which is similar to how humans process knowledge through counterfactual thinking. Thus, producing a better counterfactual map may be a step towards explanation at the level of human knowledge. For example, a counterfactual map can localize hypothetical abnormalities from a normal brain image that may cause it to be diagnosed with a disease. Specifically, our proposed BIN consists of two core components: Counterfactual Map Generator and Target Attribution Network. The Counterfactual Map Generator is a variation of conditional GAN which can synthesize a counterfactual map conditioned on an arbitrary target label. The Target Attribution Network works in a complementary manner to enforce target label attribution to the synthesized map. We have validated our proposed BIN in qualitative, quantitative analysis on MNIST, 3D Shapes, and ADNI datasets, and show the comprehensibility and fidelity of our method from various ablation studies.

Pre-trained language models (PTLM) have achieved impressive results in a range of natural language understanding (NLU) and generation (NLG) tasks. However, current pre-training objectives such as masked token prediction (for BERTstyle PTLMs) and masked span infilling (for T5-style PTLMs) do not explicitly model the relational commonsense knowledge about everyday concepts, which is crucial to many downstream tasks that need common sense to understand or generate. To augment PTLMs with concept-centric commonsense knowledge, in this paper, we propose both generative and contrastive objectives for learning common sense from the text, and use them as intermediate self-supervised learning tasks for incrementally pre-training PTLMs (before task-specific fine-tuning on downstream datasets). Furthermore, we develop a joint pre-training framework to unify generative and contrastive objectives so that they can mutually reinforce each other. We show that while only incrementally pre-trained on a relatively small corpus for a few steps, CALM outperforms baseline methods by a consistent margin and even comparable with some larger PTLMs, which suggests that CALM can serve as a general, "plug-and-play" method for improving the commonsense reasoning ability of a PTLM. Pre-trained language models (PLTMs) such as BERT (Devlin et al., 2018) and T5 (Raffel et al., 2019) have revolutionized the field of NLP, yielding impressive performance on various conventional natural language understanding (NLU) and generation (NLG) tasks. BERT and its novel variants such as RoBERTa (Liu et al., 2019) and ALBERT (Lan et al., 2019) capture syntactical and semantic knowledge mainly from the pre-training task of masked language modeling, while T5-style models such as BART (Lewis et al., 2019) instead focus on masked span infilling tasks. Though yielding better performance on many downstream tasks, these pre-training objectives, however, do not explicitly guide the models to reason with concept-centric commonsense knowledge from language, including the relation and composition of daily concepts in our lives. This leaves room for equipping current PTLMs with richer commonsense reasoning ability.

In the last few decades, building regression models for non-scalar variables, including time series, text, image, and video, has attracted increasing interests of researchers from the data analytic community. In this paper, we focus on a multivariate time series regression problem. Specifically, we aim to learn mathematical mappings from multiple chronologically measured numerical variables within a certain time interval S to multiple numerical variables of interest over time interval T. Prior arts, including the multivariate regression model, the Seq2Seq model, and the functional linear models, suffer from several limitations. The first two types of models can only handle regularly observed time series. Besides, the conventional multivariate regression models tend to be biased and inefficient, as they are incapable of encoding the temporal dependencies among observations from the same time series. The sequential learning models explicitly use the same set of parameters along time, which has negative impacts on accuracy. The function-on-function linear model in functional data analysis (a branch of statistics) is insufficient to capture complex correlations among the considered time series and suffer from underfitting easily. In this paper, we propose a general functional mapping that embraces the function-on-function linear model as a special case. We then propose a non-linear function-on-function model using the fully connected neural network to learn the mapping from data, which addresses the aforementioned concerns in the existing approaches. For the proposed model, we describe in detail the corresponding numerical implementation procedures. The effectiveness of the proposed model is demonstrated through the application to two real-world problems.

Causal inference from observation data is a core problem in many scientific fields. Here we present a general supervised deep learning framework that infers causal interactions by transforming the input vectors to an image-like representation for every pair of inputs. Given a training dataset we first construct a normalized empirical probability density distribution (NEPDF) matrix. We then train a convolutional neural network (CNN) on NEPDFs for causality predictions. We tested the method on several different simulated and real world data and compared it to prior methods for causal inference. As we show, the method is general, can efficiently handle very large datasets and improves upon prior methods.

We motivate Energy-Based Models (EBMs) as a promising model class for continual learning problems. Instead of tackling continual learning via the use of external memory, growing models, or regularization, EBMs have a natural way to support a dynamically-growing number of tasks or classes that causes less interference with previously learned information. We find that EBMs outperform the baseline methods by a large margin on several continual learning benchmarks. We also show that EBMs are adaptable to a more general continual learning setting where the data distribution changes without the notion of explicitly delineated tasks. These observations point towards EBMs as a class of models naturally inclined towards the continual learning regime.

MCUNet embeds deep learning neural networks on the off-shelf microcontrollers to reduce memory usage. Artificial Intelligence is a technology which is getting heavily researched on a routine basis. Researchers all around the world are working to make the application and implementation of AI faster and better. Over the years, humans have encountered instances where AI led to potential breakthroughs. Be it in the early detection of heart diseases or in discovering historical events, AI has come far since its inception.

The short answer to What is Artificial Intelligence is that it depends on who you ask. A layman with a fleeting understanding of technology would link it to robots. They'd say Artificial Intelligence is a terminator like-figure that can act and think on its own. An AI researcher would say that it's a set of algorithms that can produce results without having to be explicitly instructed to do so. And they would all be right. AI courses at Great Learning provide you with an overview of the current implementation scenario in various industries. With an in-depth introduction to artificial intelligence, you can easily master the basics for a better future in the course.

Machine learning is mostly about building good function approximations from data. And, when it comes to good approximations, deep learning algorithms have a great following as they are founded on principles of universal approximation. The adoption of these deep neural networks has been high in the past couple of years. But, as these systems scale, new challenges surface. It can be misclassification of an unexamined vulnerability to adversarial attack.

Deep learning neural networks are artificial intelligence systems that are being used for increasingly important decisions. Deep learning neural networks are used for tasks as varied as autonomous driving to diagnosing medical conditions. This type of network excels at recognizing patterns in large and complex datasets to help with decision-making. One big challenge is determining if the neural network is correct. Researchers at MIT and Harvard University have developed a quick way for a neural network to churn through data and provide a prediction along with the neural network's confidence level in its answer.