In this paper we study the classification problem in which we have access to easily obtainable surrogate for the true labels, namely complementary labels, which specify classes that observations do \textbf{not} belong to. For example, if one is familiar with monkeys but not meerkats, a meerkat is easily identified as not a monkey, so "monkey" is annotated to the meerkat as a complementary label. Specifically, let $Y$ and $\bar{Y}$ be the true and complementary labels, respectively. We first model the annotation of complementary labels via the transition probabilities $P(\bar{Y}=i|Y=j), i\neq j\in\{1,\cdots,c\}$, where $c$ is the number of classes. All the previous methods implicitly assume that the transition probabilities $P(\bar{Y}=i|Y=j)$ are identical, which is far from true in practice because humans are biased toward their own experience. For example, if a person is more familiar with monkey than prairie dog when providing complementary labels for meerkats, he/she is more likely to employ "monkey" as a complementary label. We therefore reason that the transition probabilities will be different. In this paper, we address three fundamental problems raised by learning with biased complementary labels. (1) How to estimate the transition probabilities? (2) How to modify the traditional loss functions and extend standard deep neural network classifiers to learn with biased complementary labels? (3) Does the classifier learned from examples with complementary labels by our proposed method converge to the optimal one learned from examples with true labels? Comprehensive experiments on MNIST, CIFAR10, CIFAR100, and Tiny ImageNet empirically validate the superiority of the proposed method to the current state-of-the-art methods with accuracy gains of over 10\%.

SGD (Stochastic Gradient Descent) is a popular algorithm for large scale optimization problems due to its low iterative cost. However, SGD can not achieve linear convergence rate as FGD (Full Gradient Descent) because of the inherent gradient variance. To attack the problem, mini-batch SGD was proposed to get a trade-off in terms of convergence rate and iteration cost. In this paper, a general CVI (Convergence-Variance Inequality) equation is presented to state formally the interaction of convergence rate and gradient variance. Then a novel algorithm named SSAG (Stochastic Stratified Average Gradient) is introduced to reduce gradient variance based on two techniques, stratified sampling and averaging over iterations that is a key idea in SAG (Stochastic Average Gradient). Furthermore, SSAG can achieve linear convergence rate of $\mathcal {O}((1-\frac{\mu}{8CL})^k)$ at smaller storage and iterative costs, where $C\geq 2$ is the category number of training data. This convergence rate depends mainly on the variance between classes, but not on the variance within the classes. In the case of $C\ll N$ ($N$ is the training data size), SSAG's convergence rate is much better than SAG's convergence rate of $\mathcal {O}((1-\frac{\mu}{8NL})^k)$. Our experimental results show SSAG outperforms SAG and many other algorithms.

In this project we analysed how much semantic information images carry, and how much value image data can add to sentiment analysis of the text associated with the images. To better understand the contribution from images, we compared models which only made use of image data, models which only made use of text data, and models which combined both data types. We also analysed if this approach could help sentiment classifiers generalize to unknown sentiments.

With the widespread use of information technologies, information networks have increasingly become popular to capture complex relationships across various disciplines, such as social networks, citation networks, telecommunication networks, and biological networks. Analyzing these networks sheds light on different aspects of social life such as the structure of society, information diffusion, and different patterns of communication. However, the large scale of information networks often makes network analytic tasks computationally expensive and intractable. Recently, network representation learning has been proposed as a new learning paradigm that embeds network vertices into a low-dimensional vector space, by preserving network topology structure, vertex content, and other side information. This facilitates the original network to be easily handled in the new vector space for further analysis. In this survey, we perform a thorough review of the current literature on network representation learning in the field of data mining and machine learning. We propose a new categorization to analyze and summarize state-of-the-art network representation learning techniques according to the methodology they employ and the network information they preserve. Finally, to facilitate research on this topic, we summarize benchmark datasets and evaluation methodologies, and discuss open issues and future research directions in this field.

We introduce neural networks for end-to-end differentiable proving of queries to knowledge bases by operating on dense vector representations of symbols. These neural networks are constructed recursively by taking inspiration from the backward chaining algorithm as used in Prolog. Specifically, we replace symbolic unification with a differentiable computation on vector representations of symbols using a radial basis function kernel, thereby combining symbolic reasoning with learning subsymbolic vector representations. By using gradient descent, the resulting neural network can be trained to infer facts from a given incomplete knowledge base. It learns to (i) place representations of similar symbols in close proximity in a vector space, (ii) make use of such similarities to prove queries, (iii) induce logical rules, and (iv) use provided and induced logical rules for multi-hop reasoning. We demonstrate that this architecture outperforms ComplEx, a state-of-the-art neural link prediction model, on three out of four benchmark knowledge bases while at the same time inducing interpretable function-free first-order logic rules.

Sequential decision making problems, such as structured prediction, robotic control, and game playing, require a combination of planning policies and generalisation of those plans. In this paper, we present Expert Iteration (ExIt), a novel reinforcement learning algorithm which decomposes the problem into separate planning and generalisation tasks. Planning new policies is performed by tree search, while a deep neural network generalises those plans. Subsequently, tree search is improved by using the neural network policy to guide search, increasing the strength of new plans. In contrast, standard deep Reinforcement Learning algorithms rely on a neural network not only to generalise plans, but to discover them too. We show that ExIt outperforms REINFORCE for training a neural network to play the board game Hex, and our final tree search agent, trained tabula rasa, defeats MoHex 1.0, the most recent Olympiad Champion player to be publicly released.

The Stanford Statistical Machine Learning Group at Stanford is a unique blend of faculty, students, and post-docs spanning AI, systems, theory, and statistics. Our work spans the spectrum from answering deep, foundational questions in the theory of machine learning to building practical large-scale machine learning algorithms which are widely used in industry. Topics include reliable machine learning, large-scale optimization, interactive learning, unsupervised and semi-supervised learning, reinforcement learning, deep learning, and statistical learning theory.

Roundup Here's a human-compiled, totally non-robot generated summary of AI news beyond what we've already reported the past month week. By the way, your humble Reg vulture will be at the machine-learning super-conference NIPS in LA next week – please do email in if you want to say hi, or point out any hot talks or gossip. AI Index – A team of AI experts have published this year's annual AI Index report that gathers data to show how the field is progressing and changing over time. The rise of deep learning has accelerated the AI hype, and it can be difficult to have meaningful conversations and shape policy without basic metrics. The report shows that the number of active AI startups has increased 14 times since 2000, and venture capital has risen six times across the same period.

In this special guest feature from Scientific Computing World, David Yip, HPC and Storage Business Development at OCF, provides his take on the place of GPU technology in HPC. There was an interesting story published earlier this week in which NVIDIA's founder and CEO, Jensen Huang, said: 'As advanced parallel-instruction architectures for CPU can be barely worked out by designers, GPUs will soon replace CPUs'. There are only so many processing cores you can fit on a single CPU chip. There are optimized applications that take advantage of a number of cores, but typically they are used for sequential serial processing (although Intel is doing an excellent job of adding more and more cores to its CPUs and getting developers to program multicore systems). By contrast, a GPU has massively parallel architecture consisting of many thousands of smaller, more efficient cores designed for handling multiple tasks simultaneously.

THAT a computer program can repeatedly beat the world champion at Go, a complex board game, is a coup for the fast-moving field of artificial intelligence (AI). Another high-stakes game, however, is taking place behind the scenes, as firms compete to hire the smartest AI experts. Technology giants, including Google, Facebook, Microsoft and Baidu, are racing to expand their AI activities. Last year they spent some $8.5 billion on deals, says Quid, a data firm. That was four times more than in 2010.