Online class-incremental continual learning (CL) studies the problem of learning new classes continually from an online non-stationary data stream, intending to adapt to new data while mitigating catastrophic forgetting. While memory replay has shown promising results, the recency bias in online learning caused by the commonly used Softmax classifier remains an unsolved challenge. Although the Nearest-Class-Mean (NCM) classifier is significantly undervalued in the CL community, we demonstrate that it is a simple yet effective substitute for the Softmax classifier. It addresses the recency bias and avoids structural changes in the fully-connected layer for new classes. Moreover, we observe considerable and consistent performance gains when replacing the Softmax classifier with the NCM classifier for several state-of-the-art replay methods. To leverage the NCM classifier more effectively, data embeddings belonging to the same class should be clustered and well-separated from those with a different class label. To this end, we contribute Supervised Contrastive Replay (SCR), which explicitly encourages samples from the same class to cluster tightly in embedding space while pushing those of different classes further apart during replay-based training. Overall, we observe that our proposed SCR substantially reduces catastrophic forgetting and outperforms state-of-the-art CL methods by a significant margin on a variety of datasets.

As data is generated and stored almost everywhere, learning a model from a data-decentralized setting is a task of interest for many AI-driven service providers. Although federated learning is settled down as the main solution in such situations, there still exists room for improvement in terms of personalization. Training federated learning systems usually focuses on optimizing a global model that is identically deployed to all client devices. However, a single global model is not sufficient for each client to be personalized on their performance as local data assumes to be not identically distributed across clients. We propose a method to address this situation through the lens of ensemble learning based on the construction of a low-loss subspace continuum that generates a high-accuracy ensemble of two endpoints (i.e. global model and local model). We demonstrate that our method achieves consistent gains both in personalized and unseen client evaluation settings through extensive experiments on several standard benchmark datasets.

Deb Norton spends her days helping teachers in Wisconsin's Oshkosh Area school district get more comfortable with technology tools they're using to engage students. A few years ago, she started seeing increasing mentions of artificial intelligence. Around then, the International Society for Technology in Education asked her to lead a course on the uses of artificial intelligence in the K-12 classroom. She was initially intrigued when she saw students light up at the mention of artificial intelligence. It soon became clear to her that they were already experiencing AI in their daily lives, with tools like Instagram filters or chatbots on websites.

Marktechpost, LLC. is a California-based Artificial Intelligence Media Platform for the latest updates in machine learning, deep learning, and data science research. Marktechpost's key focus is on spreading AI Awareness across the globe. The Marktechpost AI Introductory Course is a basic Artificial Intelligence (AI) Intro Course comprised of four video lectures. This course will cover what AI is, how it works, and why AI is taking off now. Fabio is a data scientist from Italy.

Hosted by Ben Byford, The Machine Ethics Podcast brings together interviews with academics, authors, business leaders, designers and engineers on the subject of autonomous algorithms, artificial intelligence, machine learning, and technology's impact on society. We chat about managing the risks of AI research, how the AI community should think about the consequences of their research, documenting best practises for AI, OpenAI's GTP2 research disclosure example, considering unintended consequences & negative downstream outcomes, considering what your research may actually contribute, promoting scientific openness, proportional ethical reflection, research social impact assessments and more… Madhulika Srikumar is a program lead at the Safety-Critical AI initiative at Partnership on AI, a multistakeholder non-profit shaping the future of responsible AI. Core areas of her current focus include community engagement on responsible publication norms in AI research and diversity and inclusion in AI teams. Madhu is a lawyer by training and completed her graduate studies (LL.M) at Harvard Law School. This podcast was created, and is run by, Ben Byford and collaborators.

Exploration of indoor environments has recently experienced a significant interest, also thanks to the introduction of deep neural agents built in a hierarchical fashion and trained with Deep Reinforcement Learning (DRL) on simulated environments. Current state-of-the-art methods employ a dense extrinsic reward that requires the complete a priori knowledge of the layout of the training environment to learn an effective exploration policy. However, such information is expensive to gather in terms of time and resources. In this work, we propose to train the model with a purely intrinsic reward signal to guide exploration, which is based on the impact of the robot's actions on the environment. So far, impact-based rewards have been employed for simple tasks and in procedurally generated synthetic environments with countable states. Since the number of states observable by the agent in realistic indoor environments is non-countable, we include a neural-based density model and replace the traditional count-based regularization with an estimated pseudo-count of previously visited states. The proposed exploration approach outperforms DRL-based competitors relying on intrinsic rewards and surpasses the agents trained with a dense extrinsic reward computed with the environment layouts. We also show that a robot equipped with the proposed approach seamlessly adapts to point-goal navigation and real-world deployment.

In the past few years, a considerable amount of research has been dedicated to the exploitation of previous learning experiences and the design of Few-shot and Meta Learning approaches, in problem domains ranging from Computer Vision to Reinforcement Learning based control. A notable exception, where to the best of our knowledge, little to no effort has been made in this direction is Quality-Diversity (QD) optimisation. QD methods have been shown to be effective tools in dealing with deceptive minima and sparse rewards in Reinforcement Learning. However, they remain costly due to their reliance on inherently sample inefficient evolutionary processes. We show that, given examples from a task distribution, information about the paths taken by optimisation in parameter space can be leveraged to build a prior population, which when used to initialise QD methods in unseen environments, allows for few-shot adaptation. Our proposed method does not require backpropagation. It is simple to implement and scale, and furthermore, it is agnostic to the underlying models that are being trained. Experiments carried in both sparse and dense reward settings using robotic manipulation and navigation benchmarks show that it considerably reduces the number of generations that are required for QD optimisation in these environments.

A backdoor data poisoning attack is an adversarial attack wherein the attacker injects several watermarked, mislabeled training examples into a training set. The watermark does not impact the test-time performance of the model on typical data; however, the model reliably errs on watermarked examples. To gain a better foundational understanding of backdoor data poisoning attacks, we present a formal theoretical framework within which one can discuss backdoor data poisoning attacks for classification problems. We then use this to analyze important statistical and computational issues surrounding these attacks. On the statistical front, we identify a parameter we call the memorization capacity that captures the intrinsic vulnerability of a learning problem to a backdoor attack. This allows us to argue about the robustness of several natural learning problems to backdoor attacks. Our results favoring the attacker involve presenting explicit constructions of backdoor attacks, and our robustness results show that some natural problem settings cannot yield successful backdoor attacks. From a computational standpoint, we show that under certain assumptions, adversarial training can detect the presence of backdoors in a training set. We then show that under similar assumptions, two closely related problems we call backdoor filtering and robust generalization are nearly equivalent. This implies that it is both asymptotically necessary and sufficient to design algorithms that can identify watermarked examples in the training set in order to obtain a learning algorithm that both generalizes well to unseen data and is robust to backdoors.

Benefits of chatbots 2021: To define it in simple terms, chatbots are artificial intelligence-driven (AI) robots that are integrated into smart devices so that they can talk with humans without letting them know that they are conversing with something artificial. Chatbots use natural language processing (NLP) and sentiment analysis to stimulate conversations with users. Chatbots are known by different names today. Some of these names are interactive agents, smart bots, digital assistants. Chatbots are the most innovative and significant developments in the domain of AI.

Only a few months ago, there was a brief window of time when many New Yorkers, among others, watched as the numbers of the vaccinated climbed and dared to hope that the year-long pandemic was finally coming to an end. Vacations were booked, weddings were scheduled, and parents began looking forward to getting their children out of the living room and back to attending school in person. But, as Barry Blitt captures in his new cover, the pandemic has not gone away, and, for students and their parents, the usual anxieties around returning to the classroom have been compounded by an increasing incidence of coronavirus infections in children, many of whom are too young to be vaccinated, and other related uncertainties. We recently spoke to Blitt about back-to-school blues and presenting his work at elementary schools. Were you a good student?