We consider interactive learning in the realizable setting and develop a general framework to handle problems ranging from best arm identification to active classification. We begin our investigation with the observation that agnostic algorithms \emph{cannot} be minimax-optimal in the realizable setting. Hence, we design novel computationally efficient algorithms for the realizable setting that match the minimax lower bound up to logarithmic factors and are general-purpose, accommodating a wide variety of function classes including kernel methods, H{\"o}lder smooth functions, and convex functions. The sample complexities of our algorithms can be quantified in terms of well-known quantities like the extended teaching dimension and haystack dimension. However, unlike algorithms based directly on those combinatorial quantities, our algorithms are computationally efficient. To achieve computational efficiency, our algorithms sample from the version space using Monte Carlo "hit-and-run" algorithms instead of maintaining the version space explicitly. Our approach has two key strengths. First, it is simple, consisting of two unifying, greedy algorithms. Second, our algorithms have the capability to seamlessly leverage prior knowledge that is often available and useful in practice. In addition to our new theoretical results, we demonstrate empirically that our algorithms are competitive with Gaussian process UCB methods.

The equivalence of realizable and agnostic learnability is a fundamental phenomenon in learning theory. With variants ranging from classical settings like PAC learning and regression to recent trends such as adversarially robust and private learning, it's surprising that we still lack a unified theory; traditional proofs of the equivalence tend to be disparate, and rely on strong model-specific assumptions like uniform convergence and sample compression. In this work, we give the first model-independent framework explaining the equivalence of realizable and agnostic learnability: a three-line blackbox reduction that simplifies, unifies, and extends our understanding across a wide variety of settings. This includes models with no known characterization of learnability such as learning with arbitrary distributional assumptions or general loss, as well as a host of other popular settings such as robust learning, partial learning, fair learning, and the statistical query model. More generally, we argue that the equivalence of realizable and agnostic learning is actually a special case of a broader phenomenon we call property generalization: any desirable property of a learning algorithm (e.g.\ noise tolerance, privacy, stability) that can be satisfied over finite hypothesis classes extends (possibly in some variation) to any learnable hypothesis class.

Artificial Intelligence has been transforming industries and academic research across the globe, and research software development is no exception. Machine learning and deep learning are being applied in every aspect of the research software development lifecycles, from new algorithm design paradigms to software development processes. In this paper, we discuss our views on today's challenges and opportunities that AI has presented on research software development and engineers, and the approaches we, at the University of Florida, are taking to prepare our workforce for the new era of AI.

In real world, affecting the environment by a weak policy can be expensive or very risky, therefore hampers real world applications of reinforcement learning. Offline Reinforcement Learning (RL) can learn policies from a given dataset without interacting with the environment. However, the dataset is the only source of information for an Offline RL algorithm and determines the performance of the learned policy. We still lack studies on how dataset characteristics influence different Offline RL algorithms. Therefore, we conducted a comprehensive empirical analysis of how dataset characteristics effect the performance of Offline RL algorithms for discrete action environments. A dataset is characterized by two metrics: (1) the average dataset return measured by the Trajectory Quality (TQ) and (2) the coverage measured by the State-Action Coverage (SACo). We found that variants of the off-policy Deep Q-Network family require datasets with high SACo to perform well. Algorithms that constrain the learned policy towards the given dataset perform well for datasets with high TQ or SACo. For datasets with high TQ, Behavior Cloning outperforms or performs similarly to the best Offline RL algorithms.

During my studies at Imperial College London - 3ish years ago - I was introduced to the topic of Machine Learning, an area that I had always been interested about. At that time, not only I didn't know python, as "from sklearn.decomposition Therefore, we had to write the PCA (and other) algorithms in Matlab from scratch, which was great because it exposed us to the maths behind each algorithm.

Exposing children to AI helps them build logical reasoning skills and helps them understand use of digital tools with higher awareness.

The dramatic rise of online learning during the COVID-19 pandemic has spotlit concerns about the role of technology in exam surveillance -- and also in student cheating. Some universities have reported more cheating during the pandemic, and such concerns are unfolding in a climate where technologies that allow for the automation of writing continue to improve. Over the past two years, the ability of artificial intelligence to generate writing has leapt forward significantly, particularly with the development of what's known as the language generator GPT-3. With this, companies such as Google, Microsoft and NVIDIA can now produce "human-like" text. AI-generated writing has raised the stakes of how universities and schools will gauge what constitutes academic misconduct, such as plagiarism.

Thanks to rapid development of artificial intelligence technology in recent years, the current artificial intelligence technology is contributing to many part of society. Education, environment, medical care, military, tourism, economy, politics, etc. are having a very large impact on society as a whole. For example, in the field of education, there is an artificial intelligence tutoring system that automatically assigns tutors based on student's level. In the field of economics, there are quantitative investment methods that automatically analyze large amounts of data to find investment laws to create investment models or predict changes in financial markets. As such, artificial intelligence technology is being used in various fields. So, it is very important to know exactly what factors have an important influence on each field of artificial intelligence technology and how the relationship between each field is connected. Therefore, it is necessary to analyze artificial intelligence technology in each field. In this paper, we analyze patent documents related to artificial intelligence technology. We propose a method for keyword analysis within factors using artificial intelligence patent data sets for artificial intelligence technology analysis. This is a model that relies on feature engineering based on deep learning model named KeyBERT, and using vector space model. A case study of collecting and analyzing artificial intelligence patent data was conducted to show how the proposed model can be applied to real world problems.

Quantile (and, more generally, KL) regret bounds, such as those achieved by NormalHedge (Chaudhuri, Freund, and Hsu 2009) and its variants, relax the goal of competing against the best individual expert to only competing against a majority of experts on adversarial data. More recently, the semi-adversarial paradigm (Bilodeau, Negrea, and Roy 2020) provides an alternative relaxation of adversarial online learning by considering data that may be neither fully adversarial nor stochastic (i.i.d.). We achieve the minimax optimal regret in both paradigms using FTRL with separate, novel, root-logarithmic regularizers, both of which can be interpreted as yielding variants of NormalHedge. We extend existing KL regret upper bounds, which hold uniformly over target distributions, to possibly uncountable expert classes with arbitrary priors; provide the first full-information lower bounds for quantile regret on finite expert classes (which are tight); and provide an adaptively minimax optimal algorithm for the semi-adversarial paradigm that adapts to the true, unknown constraint faster, leading to uniformly improved regret bounds over existing methods.

Reinforcement Learning (RL) agents can learn to solve complex sequential decision making tasks by interacting with the environment. However, sample efficiency remains a major challenge. In the field of multi-goal RL, where agents are required to reach multiple goals to solve complex tasks, improving sample efficiency can be especially challenging. On the other hand, humans or other biological agents learn such tasks in a much more strategic way, following a curriculum where tasks are sampled with increasing difficulty level in order to make gradual and efficient learning progress. In this work, we propose a method for automatic goal generation using a dynamical distance function (DDF) in a self-supervised fashion. DDF is a function which predicts the dynamical distance between any two states within a markov decision process (MDP). With this, we generate a curriculum of goals at the appropriate difficulty level to facilitate efficient learning throughout the training process. We evaluate this approach on several goal-conditioned robotic manipulation and navigation tasks, and show improvements in sample efficiency over a baseline method which only uses random goal sampling.