Search for: All records

n modern supervised learning, there are a large number of tasks, but many of them are associated with only a small amount of labelled data. These include data from medical image processing and robotic interaction. Even though each individual task cannot be meaningfully trained in isolation, one seeks to meta-learn across the tasks from past experiences by exploiting some similarities. We study a fundamental question of interest: When can abundant tasks with small data compensate for lack of tasks with big data? We focus on a canonical scenario where each task is drawn from a mixture of k linear regressions, and identify sufficient conditions for such a graceful exchange to hold; there is little loss in sample complexity even when we only have access to small data tasks. To this end, we introduce a novel spectral approach and show that we can efficiently utilize small data tasks with the help of Omega(k^3/2) medium data tasks each with Omega(k^1/2) examples. 

Societal-scale data is playing an increasingly prominent role in social science research; examples from research on geopolitical events include questions on how emergency events impact the diffusion of information or how new policies change patterns of social interaction. Such research often draws critical inferences from observing how an exogenous event changes meaningful metrics like network degree or network entropy. However, as we show in this work, standard estimation methodologies make systematically incorrect inferences when the event also changes the sparsity of the data. To address this issue, we provide a general framework for inferring changes in social metrics when dealing with non-stationary sparsity. We propose a plug-in correction that can be applied to any estimator, including several recently proposed procedures. Using both simulated and real data, we demonstrate that the correction significantly improves the accuracy of the estimated change under a variety of plausible data generating processes. In particular, using a large dataset of calls from Afghanistan, we show that whereas traditional methods substantially overestimate the impact of a violent event on social diversity, the plug-in correction reveals the true response to be much more modest. 

With the rapid expansion of mobile phone networks in developing countries, large-scale graph machine learning has gained sudden relevance in the study of global poverty. Recent applications range from humanitarian response and poverty estimation to urban planning and epidemic containment. Yet the vast majority of computational tools and algorithms used in these applications do not account for the multi-view nature of social networks: people are related in myriad ways, but most graph learning models treat relations as binary. In this paper, we develop a graph-based convolutional network for learning on multi-view networks. We show that this method outperforms state-of-the-art semi-supervised learning algorithms on three different prediction tasks using mobile phone datasets from three different developing countries. We also show that, while designed specifically for use in poverty research, the algorithm also outperforms existing benchmarks on a broader set of learning tasks on multi-view networks, including node labelling in citation networks. 

Consider a setting with N independent individuals, each with an unknown parameter, p_i in [0,1] drawn from some unknown distribution P*. After observing the outcomes of t independent Bernoulli trials, i.e., Xi ~ Binomial(t, p_i) per individual, our objective is to accurately estimate P*. This problem arises in numerous domains, including the social sciences, psychology, healthcare, and biology, where the size of the population under study is usually large while the number of observations per individual is often limited. Our main result shows that, in the regime where t << N , the maximum likelihood estimator (MLE) is both statisticallyminimax optimal and efficiently computable. Precisely, for sufficiently large N , the MLE achieves the information theoretic optimal error bound of O(1/sqrt(t log N)) for N< exp(t), and O(1/t) for N> exp(t), with regards to the L1 distance between the true cdf and the estimated cdf. 

We consider the problem of accurately recovering a matrix B of size M by M, which represents a probability distribution over M^2 outcomes, given access to an observed matrix of "counts" generated by taking independent samples from the distribution B. How can structural properties of the underlying matrix B be leveraged to yield computationally efficient and information theoretically optimal reconstruction algorithms? When can accurate reconstruction be accomplished in the sparse data regime? This basic problem lies at the core of a number of questions that are currently being considered by different communities, including building recommendation systems and collaborative filtering in the sparse data regime, community detection in sparse random graphs, learning structured models such as topic models or hidden Markov models, and the efforts from the natural language processing community to compute "word embeddings". Many aspects of this problem---both in terms of learning and property testing/estimation and on both the algorithmic and information theoretic sides---remain open. Our results apply to the setting where B has a low rank structure. For this setting, we propose an efficient (and practically viable) algorithm that accurately recovers the underlying M by M matrix using O(M) samples} (where we assume the rank is a constant). This linear sample complexity is optimal, up to constant factors, in an extremely strong sense: even testing basic properties of the underlying matrix (such as whether it has rank 1 or 2) requires Omega(M) samples. Additionally, we provide an even stronger lower bound showing that distinguishing whether a sequence of observations were drawn from the uniform distribution over M observations versus being generated by a well-conditioned Hidden Markov Model with two hidden states requires Omega(M) observations, while our positive results for recovering B immediately imply that Omega(M) observations suffice to learn such an HMM. This lower bound precludes sublinear-sample hypothesis tests for basic properties, such as identity or uniformity, as well as sublinear sample estimators for quantities such as the entropy rate of HMMs. 

Historically, economists have relied heavily on survey-based data collection to measure social and economic well-being. Here, we investigate the extent to which the "digital footprints" of an individual can be used to infer his or her socioeconomic characteristics. Using two different datasets from Afghanistan and Rwanda, we show that phone data can be used to estimate the wealth of individuals in two very different economic environments. However, we find that such models are relatively brittle, and that a model trained in one country cannot be used to estimate characteristics in another. These results suggest several promising applications and directions for future work.