I want to tell you straight off what this story is about: Sometime in the next 40 years, robots are going to take your job. I don't care what your job is. If you dig ditches, a robot will dig them better. If you're a magazine writer, a robot will write your articles better. If you're a doctor, IBM's Watson will no longer "assist" you in finding the right diagnosis from its database of millions of case studies and journal articles. It will just be a better doctor than you. Robots will run companies better than you do. Robots will paint and write and sculpt better than you. Think you have social skills that no robot can match? Within 20 years, maybe half of you will be out of jobs. A couple of decades after that, most of the rest of you will be out of jobs. In one sense, this all sounds great. Let the robots have the damn jobs! We'll be free to read or write poetry or play video games or whatever we want to do.

A survey of 500 CIOs from around the world by ServiceNow finds that machine learning has arrived in the enterprise, making material contributions to everyday work. To realize its full value, technology leaders must find skilled talent to work side-by-side with machines, in addition to redesigning their organizations and processes. Go from Pro to Superhero! Our Amazing, Incredible, Invincible PR Strategy Checklist is jam-packed with wisdom and resources to keep you saving the day! For The Global CIO Point of View, ServiceNow surveyed CIOs in 11 countries across 25 industries to uncover the competitive benefits of adopting machine learning and hear how those leaders are driving results.

If my email inbox is anything to go by, a technology revolution is under way that is going to transform all of our lives very soon and it is called artificial intelligence. A Welsh company is using AI to detect North Korean bio-weapons. I could pop over to California to hear about "AI wearable solutions for aging population". And Lloyd's of London has unveiled an artificial intelligence partnership with a firm that promises "in a decade a significant part of the insurance industry will be powered by AI". These represent just three of the innumerable AI press releases aimed at me and other technology journalists over recent days.

In this paper, we provide an approach to clustering relational matrices whose entries correspond to either similarities or dissimilarities between objects. Our approach is based on the value of information, a parameterized, information-theoretic criterion that measures the change in costs associated with changes in information. Optimizing the value of information yields a deterministic annealing style of clustering with many benefits. For instance, investigators avoid needing to a priori specify the number of clusters, as the partitions naturally undergo phase changes, during the annealing process, whereby the number of clusters changes in a data-driven fashion. The global-best partition can also often be identified.

State-of-the-art methods in convex and non-convex optimization employ higher-order derivative information, either implicitly or explicitly. We explore the limitations of higher-order optimization and prove that even for convex optimization, a polynomial dependence on the approximation guarantee and higher-order smoothness parameters is necessary. As a special case, we show Nesterov's accelerated cubic regularization method to be nearly tight.

The goal of regression and classification methods in supervised learning is to minimize the empirical risk, that is, the expectation of some loss function quantifying the prediction error under the empirical distribution. When facing scarce training data, overfitting is typically mitigated by adding regularization terms to the objective that penalize hypothesis complexity. In this paper we introduce new regularization techniques using ideas from distributionally robust optimization, and we give new probabilistic interpretations to existing techniques. Specifically, we propose to minimize the worst-case expected loss, where the worst case is taken over the ball of all (continuous or discrete) distributions that have a bounded transportation distance from the (discrete) empirical distribution. By choosing the radius of this ball judiciously, we can guarantee that the worst-case expected loss provides an upper confidence bound on the loss on test data, thus offering new generalization bounds. We prove that the resulting regularized learning problems are tractable and can be tractably kernelized for many popular loss functions. We validate our theoretical out-of-sample guarantees through simulated and empirical experiments.

The correlation length-scale next to the noise variance are the most used hyperparameters for the Gaussian processes. Typically, stationary covariance functions are used, which are only dependent on the distances between input points and thus invariant to the translations in the input space. The optimization of the hyperparameters is commonly done by maximizing the log marginal likelihood. This works quite well, if the distances are uniform distributed. In the case of a locally adapted or even sparse input space, the prediction of a test point can be worse dependent of its position. A possible solution to this, is the usage of a non-stationary covariance function, where the hyperparameters are calculated by a deep neural network. So that the correlation length scales and possibly the noise variance are dependent on the test point. Furthermore, different types of covariance functions are trained simultaneously, so that the Gaussian process prediction is an additive overlay of different covariance matrices. The right covariance functions combination and its hyperparameters are learned by the deep neural network. Additional, the Gaussian process will be able to be trained by batches or online and so it can handle arbitrarily large data sets. We call this framework Deep Gaussian Covariance Network (DGCP). There are also further extensions to this framework possible, for example sequentially dependent problems like time series or the local mixture of experts. The basic framework and some extension possibilities will be presented in this work. Moreover, a comparison to some recent state of the art surrogate model methods will be performed, also for a time dependent problem.

Reward augmented maximum likelihood (RAML), a simple and effective learning framework to directly optimize towards the reward function in structured prediction tasks, has led to a number of impressive empirical successes. RAML incorporates task-specific reward by performing maximum-likelihood updates on candidate outputs sampled according to an exponentiated payoff distribution, which gives higher probabilities to candidates that are close to the reference output. While RAML is notable for its simplicity, efficiency, and its impressive empirical successes, the theoretical properties of RAML, especially the behavior of the exponentiated payoff distribution, has not been examined thoroughly. In this work, we introduce softmax Q-distribution estimation, a novel theoretical interpretation of RAML, which reveals the relation between RAML and Bayesian decision theory. The softmax Q-distribution can be regarded as a smooth approximation of the Bayes decision boundary, and the Bayes decision rule is achieved by decoding with this Q-distribution. We further show that RAML is equivalent to approximately estimating the softmax Q-distribution, with the temperature $\tau$ controlling approximation error. We perform two experiments, one on synthetic data of multi-class classification and one on real data of image captioning, to demonstrate the relationship between RAML and the proposed softmax Q-distribution estimation method, verifying our theoretical analysis. Additional experiments on three structured prediction tasks with rewards defined on sequential (named entity recognition), tree-based (dependency parsing) and irregular (machine translation) structures show notable improvements over maximum likelihood baselines.

ABSTRACT Long short-term memory (LSTM) is normally used in recurrent neural network (RNN) as basic recurrent unit. However, conventional LSTM assumes that the state at current time step depends on previous time step. In this study, we propose a new variation of LSTM, advanced LSTM (A-LSTM), for better temporal context modeling. We employ A-LSTM in weighted pooling RNN for emotion recognition. The A-LSTM outperforms the conventional LSTM by 5.5% relatively. The A-LSTM based weighted pooling RNN can also complement the state-of-the-art emotion classification framework. This shows the advantage of A-LSTM. Index Terms-- multi-task learning, attention model, long short-term memory, recurrent neural network, emotion recognition 1. INTRODUCTION Recurrent neural network is recently used as a dynamic model for sequential input.

We study the notion of robustness in stable matching problems. We first define robustness by introducing (a,b)-supermatches. An (a, b)-supermatch is a stable matching in which if any a pairs break up it is possible to find another stable matching by changing the partners of those a pairs and the partners of at most b other pairs. In this context, we define the most robust stable matching as a (1, b)- supermatch where b is minimum. We first show that checking whether a given stable matching is a (1, b)-supermatch can be done in polynomial time. Next, we use this procedure to design a constraint programming model, a local search approach, and a genetic algorithm to find the most robust stable matching. Our empirical evaluation on large instances shows that local search outperforms the other approaches.