The package implements an efficient greedy selection algorithm that offers strong theoretical guarantees on the quality of the selected set. Two submodular set functions are implemented in apricot: facility location, which is broadly applicable but requires memory quadratic in the number of examples in the data set, and a feature-based function that is less broadly applicable but can scale to millions of examples. Apricot is extremely efficient, using both algorithmic speedups such as the lazy greedy algorithm and code optimizers such as numba. We demonstrate the use of subset selection by training machine learning models to comparable accuracy using either the full data set or a representative subset thereof. This paper presents an explanation of submodular selection, an overview of the features in apricot, and an application to several data sets.

Deep neural networks have achieved great success in classification tasks during the last years. However, one major problem to the path towards artificial intelligence is the inability of neural networks to accurately detect novel class distributions and therefore, most of the classification algorithms proposed make the assumption that all classes are known prior to the training stage. In this work, we propose a methodology for training a neural network that allows it to efficiently detect novel class distributions without compromising much of its classification accuracy on the test examples of known classes. Experimental results on the CIFAR 100 and MiniImagenet data sets demonstrate the effectiveness of the proposed algorithm. The way this method was constructed also makes it suitable for training any classification algorithm that is based on Maximum Likelihood methods.

In this paper, we aim to solve for unsupervised domain adaptation of classifiers where we have access to label information for the source domain while these are not available for a target domain. While various methods have been proposed for solving these including adversarial discriminator based methods, most approaches have focused on the entire image based domain adaptation. In an image, there would be regions that can be adapted better, for instance, the foreground object may be similar in nature. To obtain such regions, we propose methods that consider the probabilistic certainty estimate of various regions and specify focus on these during classification for adaptation. We observe that just by incorporating the probabilistic certainty of the discriminator while training the classifier, we are able to obtain state of the art results on various datasets as compared against all the recent methods. We provide a thorough empirical analysis of the method by providing ablation analysis, statistical significance test, and visualization of the attention maps and t-SNE embeddings. These evaluations convincingly demonstrate the effectiveness of the proposed approach.

Deep neural networks obtain state-of-the-art performance on a series of tasks. However, they are easily fooled by adding a small adversarial perturbation to input. The perturbation is often human imperceptible on image data. We observe a significant difference in feature attributions of adversarially crafted examples from those of original ones. Based on this observation, we introduce a new framework to detect adversarial examples through thresholding a scale estimate of feature attribution scores. Furthermore, we extend our method to include multi-layer feature attributions in order to tackle the attacks with mixed confidence levels. Through vast experiments, our method achieves superior performances in distinguishing adversarial examples from popular attack methods on a variety of real data sets among state-of-the-art detection methods. In particular, our method is able to detect adversarial examples of mixed confidence levels, and transfer between different attacking methods.

In this paper, we study a challenging problem of how to pool multiple ride-share trip requests in real time under an uncertain environment. The goals are better performance metrics of efficiency and acceptable satisfaction of riders. To solve the problem effectively, an objective function that compromises the benefits and losses of dynamic ridesharing service is proposed. The Polar Coordinates based Ride-Matching strategy (PCRM) that can adapt to the satisfaction of riders on board is also addressed. In the experiment, large scale data sets from New York City (NYC) are applied. We do a case study to identify the best set of parameters of the dynamic ridesharing service with a training set of 135,252 trip requests. In addition, we also use a testing set containing 427,799 trip requests and two state-of-the-art approaches as baselines to estimate the effectiveness of our method. The experimental results show that on average 38% of traveling distance can be saved, nearly 100% of passengers can be served and each rider only spends an additional 3.8 minutes in ridesharing trips compared to single rider service.

Motivated by the many real-world applications of reinforcement learning (RL) that require safe-policy iterations, we consider the problem of off-policy evaluation (OPE) --- the problem of evaluating a new policy using the historical data obtained by different behavior policies --- under the model of nonstationary episodic Markov Decision Processes with a long horizon and large action space. Existing importance sampling (IS) methods often suffer from large variance that depends exponentially on the RL horizon $H$. To solve this problem, we consider a marginalized importance sampling (MIS) estimator that recursively estimates the state marginal distribution for the target policy at every step. MIS achieves a mean-squared error of $O(H^2R_{\max}^2\sum_{t=1}^H\mathbb E_\mu[(w_{\pi,\mu}(s_t,a_t))^2]/n)$ for large $n$, where $w_{\pi,\mu}(s_t,a_t)$ is the ratio of the marginal distribution of $t$th step under $\pi$ and $\mu$, $H$ is the horizon, $R_{\max}$ is the maximal rewards, and $n$ is the sample size. The result nearly matches the Cramer-Rao lower bounds for DAG MDP in \citet{jiang2016doubly} for most non-trivial regimes. To the best of our knowledge, this is the first OPE estimator with provably optimal dependence in $H$ and the second moments of the importance weight. Besides theoretical optimality, we empirically demonstrate the superiority of our method in time-varying, partially observable, and long-horizon RL environments.

An overwhelming majority (94 per cent) of UK-based financial services industry decision-makers believe that Artificial Intelligence (AI) has the most potential to revolutionise the sector over the next five years, considerably ahead of blockchain (53 per cent) and the Internet of Things (24 per cent). Intertrust, a provider of administrative services to clients operating and investing in the international business environment, surveyed UK-based executives covering the asset management, capital markets and private wealth sectors to identify the value-add delivered by new technologies now and in the future. Some 83 per cent of respondents believe that operations roles are the most likely to be replaced or dramatically changed by AI, robotics and blockchain, ahead of accounting (82 per cent), and compliance (65 per cent). While AI is seen as the clear frontrunner among disruptive technologies, its rollout is being hindered by a skills shortage. Some 41 per cent of respondents said they are struggling to recruit AI specialists, ahead of those working in data-analytics (29 per cent), cybersecurity (18 per cent) and compliance (18 per cent).

More people die in construction than in any other industry, and the number one cause of death on a job site is falling. Autodesk's latest addition to its BIM 360 suite of artificial intelligence (AI) enabled industry tools – Construction IQ – aims to reduce these tragic occurrences. It does this by predicting when falls are likely to happen – as well as any other danger to life, limb, or even just quality of work. Autodesk's data scientists hit upon the solution while looking for applications where the massive amount of data collected on modern-day construction sites could be put to use, thanks to the industry's enthusiastic adoption of mobile tools and sensing devices. "Imagine being a construction manager and having to contend with the fact that every X number of months, someone's going to die on the job – it's unfathomable to most of us in white collar jobs," says Pat Keaney, Autodesk's lead on the Construction IQ project.

A group of Australian researchers have recently developed an AI-powered smartphone app that can diagnose respiratory disorders by "listening" to the user's cough. This technology was developed by researchers at Curtin University and The University of Queensland, Australia, whose findings were published June 6 in the journal Respiratory Research. The researchers created an algorithm that can analyze coughs for features that are unique to five different diseases. This technique is similar to speech recognition technologies in that the software examines the auditory cough for characteristics specific to these conditions. This is typically done by a physician during a clinical exam, with a stethoscope being used to listen to sound produced while breathing or coughing (auscultation). The downside to this is that the patient must be in the presence of a trained professional to have their respiration sounds analyzed.

Lately it seems that AI is our knight in shining armor, the missing link between our past and our future. AI is now being used to steer the direction of hedge funds and drive much needed efficiency upgrades to our supply chains, for example. The machine learning society introduced a face recognition algorithm that is able to distinguish gender with an accuracy of up to 91%. AI is becoming so commonplace, in fact, that our electronics are often using it in the background to improve our experience, from taking a picture or securing our devices, without us even knowing it. Smaller AI startups are introducing AI services that can help sales teams surface talking points that closes deals. According to Element AI, a specialty lab in Montreal, "in the entire world, fewer than 10,000 people have the skills necessary to tackle serious artificial intelligence research."