We study the problem of causal structure learning when the experimenter is limited to perform at most $k$ non-adaptive experiments of size $1$. We formulate the problem of finding the best intervention target set as an optimization problem, which aims to maximize the average number of edges whose directions are resolved. We prove that the objective function is submodular and a greedy algorithm is a $(1-\frac{1}{e})$-approximation algorithm for the problem. We further present an accelerated variant of the greedy algorithm, which can lead to orders of magnitude performance speedup. We validate our proposed approach on synthetic and real graphs. The results show that compared to the purely observational setting, our algorithm orients majority of the edges through only a small number of interventions.

We consider the problem of recovering a $d-$dimensional manifold $\mathcal{M} \subset \mathbb{R}^n$ when provided with noiseless samples from $\mathcal{M}$. There are many algorithms (e.g., Isomap) that are used in practice to fit manifolds and thus reduce the dimensionality of a given data set. Ideally, the estimate $\mathcal{M}_\mathrm{put}$ of $\mathcal{M}$ should be an actual manifold of a certain smoothness; furthermore, $\mathcal{M}_\mathrm{put}$ should be arbitrarily close to $\mathcal{M}$ in Hausdorff distance given a large enough sample. Generally speaking, existing manifold learning algorithms do not meet these criteria. Fefferman, Mitter, and Narayanan (2016) have developed an algorithm whose output is provably a manifold. The key idea is to define an approximate squared-distance function (asdf) to $\mathcal{M}$. Then, $\mathcal{M}_\mathrm{put}$ is given by the set of points where the gradient of the asdf is orthogonal to the subspace spanned by the largest $n - d$ eigenvectors of the Hessian of the asdf. As long as the asdf meets certain regularity conditions, $\mathcal{M}_\mathrm{put}$ is a manifold that is arbitrarily close in Hausdorff distance to $\mathcal{M}$. In this paper, we define two asdfs that can be calculated from the data and show that they meet the required regularity conditions. The first asdf is based on kernel density estimation, and the second is based on estimation of tangent spaces using local principal components analysis.

Various measures can be used to estimate bias or unfairness in a predictor. Previous work has already established that some of these measures are incompatible with each other. Here we show that, when groups differ in prevalence of the predicted event, several intuitive, reasonable measures of fairness (probability of positive prediction given occurrence or non-occurrence; probability of occurrence given prediction or non-prediction; and ratio of predictions over occurrences for each group) are all mutually exclusive: if one of them is equal among groups, the other two must differ. The only exceptions are for perfect, or trivial (always-positive or always-negative) predictors. As a consequence, any non-perfect, non-trivial predictor must necessarily be "unfair" under two out of three reasonable sets of criteria. This result readily generalizes to a wide range of well-known statistical quantities (sensitivity, specificity, false positive rate, precision, etc.), all of which can be divided into three mutually exclusive groups. Importantly, The results applies to all predictors, whether algorithmic or human. We conclude with possible ways to handle this effect when assessing and designing prediction methods.

These and many other fascinating insights are from the Boston Consulting Group and MIT Sloan Management Review study published this week, Reshaping Business With Artificial Intelligence. An online summary of the report is available, and a PDF of the report is accessible here (22 pp., PDF, free, no opt-in). The survey is based on interviews with more than 3,000 business executives, managers, and analysts in 112 countries and 21 industries. For additional details regarding the methodology, please see page 4. The research found significant gaps between companies who have already adopted and understand Artificial Intelligence (AI) and those lagging. AI early adopters invest heavily in analytics expertise and ensuring the quality of algorithms and data can scale across their enterprise-wide information and knowledge needs.

Natural Language Processing, or NLP, is a subfield of machine learning concerned with understanding speech and text data. Statistical methods and statistical machine learning dominate the field and more recently deep learning methods have proven very effective in challenging NLP problems like speech recognition and text translation. In this post, you will discover the Stanford course on the topic of Natural Language Processing with Deep Learning methods. This course is free and I encourage you to make use of this excellent resource. The course is taught by Chris Manning and Richard Socher.

Creating noodling piano tunes and endless configurations of cat drawings with AI may not sound like an obvious project for Google, but it makes a lot of sense to Douglas Eck. Eck has spent about 15 years studying AI and music, and these days he's a research scientist on the Google Brain team, leading Magenta--Google's open-source research project that's aimed at making art and music with machine learning. He spoke to MIT Technology Review about how Google is producing new sounds with deep neural networks, where Magenta is taking AI music, and why computers suck at telling jokes. Below is an edited excerpt of the interview. Premium MIT Technology Review subscribers can listen to the full interview. Using AI to make art isn't new, so what's unique about Google's approach?

Machine learning is easily one of the biggest buzzwords in tech right now. Over the past three years Google searches for "machine learning" have increased by over 350%. But understanding machine learning can be difficult -- you either use pre-built packages that act like'black boxes' where you pass in data and magic comes out the other end, or you have to deal with high level maths and linear algebra. This tutorial is designed to introduce you to the fundamental concepts of machine learning -- you'll build your very first model from scratch to make predictions, while understanding exactly how your model works. This tutorial is based on our Dataquest Machine Learning Fundamentals course, which is part of our Data Science Learning Path. The course goes into a lot more detail, and allows you to follow along writing code to learn by doing.

Voice assistants, including Apple's Siri and Amazon's Alexa, can be controlled by hackers using inaudible voice commands, researchers at Zhejiang University in China have found. This can be done using a technique that translates voice commands into ultrasonic frequencies that are too high for the human ear to recognise. The technique, named DolphinAttack, could be used to download a virus, send fake messages and even add fake events to a calendar. It could also give hackers access to outgoing video or phone calls, allowing them to spy on their victims. The fault is due to vulnerabilities in the software and hardware of speech recognition systems.

Are you interested in data science and machine learning? If yes, consider looking at the courses we list in this post. These online classes will help you get a head start in this field. You can start and stop at any time, no hard and fast rules at all. The best part is, these courses are often available for just $10 to $15. Content: The course consists of 277 lectures, a total of 40.5 hours of video lessons.

The Long Short-Term Memory recurrent neural network has the promise of learning long sequences of observations. It seems a perfect match for time series forecasting, and in fact, it may be. In this tutorial, you will discover how to develop an LSTM forecast model for a one-step univariate time series forecasting problem. Time Series Forecasting with the Long Short-Term Memory Network in Python Photo by Matt MacGillivray, some rights reserved. This is a big topic and we are going to cover a lot of ground. This tutorial assumes you have a Python SciPy environment installed. You can use either Python 2 or 3 with this tutorial. You must have Keras (2.0 or higher) installed with either the TensorFlow or Theano backend.