We propose in this paper an exploratory analysis algorithm for functional data. The method partitions a set of functions into $K$ clusters and represents each cluster by a simple prototype (e.g., piecewise constant). The total number of segments in the prototypes, $P$, is chosen by the user and optimally distributed among the clusters via two dynamic programming algorithms. The practical relevance of the method is shown on two real world datasets.

The class of Schoenberg transformations, embedding Euclidean distances into higher dimensional Euclidean spaces, is presented, and derived from theorems on positive definite and conditionally negative definite matrices. Original results on the arc lengths, angles and curvature of the transformations are proposed, and visualized on artificial data sets by classical multidimensional scaling. A simple distance-based discriminant algorithm illustrates the theory, intimately connected to the Gaussian kernels of Machine Learning.

We present a novel framework for learning to interpret and generate language using only perceptual context as supervision. We demonstrate its capabilities by developing a system that learns to sportscast simulated robot soccer games in both English and Korean without any language-specific prior knowledge. Training employs only ambiguous supervision consisting of a stream of descriptive textual comments and a sequence of events extracted from the simulation trace. The system simultaneously establishes correspondences between individual comments and the events that they describe while building a translation model that supports both parsing and generation. We also present a novel algorithm for learning which events are worth describing. Human evaluations of the generated commentaries indicate they are of reasonable quality and in some cases even on par with those produced by humans for our limited domain.

Probabilistic matrix factorization (PMF) is a powerful method for modeling data associated with pairwise relationships, finding use in collaborative filtering, computational biology, and document analysis, among other areas. In many domains, there is additional information that can assist in prediction. For example, when modeling movie ratings, we might know when the rating occurred, where the user lives, or what actors appear in the movie. It is difficult, however, to incorporate this side information into the PMF model. We propose a framework for incorporating side information by coupling together multiple PMF problems via Gaussian process priors. We replace scalar latent features with functions that vary over the space of side information. The GP priors on these functions require them to vary smoothly and share information. We successfully use this new method to predict the scores of professional basketball games, where side information about the venue and date of the game are relevant for the outcome.

But whereas in the general Rwandan populace males tend Despite the increasing ubiquity of mobile phones in the developing to be much better educated (76.3% of males are literate, but world, remarkably little is known about the structure only 64.7% of females), among mobile phone users it is the and demographics of the mobile phone market. While a women who achieve higher levels of education: the median few qualitative studies have detailed social norms of phone woman completes secondary school, while the median man use in specific communities (Donner 2007; Burrell 2009), does not (t 4.79). Table 1 shows a few statistics on asset and a handful of quantitative researchers have begun to analyze ownership, with associated sampling error.

Diabetic retinopathy, an eye disorder caused by diabetes, is the primary cause of blindness in America and over 99% of cases in India. India and China currently account for over 90 million diabetic patients and are on the verge of an explosion of diabetic populations. This may result in an unprecedented number of persons becoming blind unless diabetic retinopathy can be detected early. Aravind Eye Hospitals is the largest eye care facility in the world, handling over 2 million patients per year. The hospital is on a massive drive throughout southern India to detect diabetic retinopathy at an early stage. To that end, a group of 10-15 physicians are responsible for manually diagnosing over 2 million retinal images per year to detect diabetic retinopathy. While the task is extremely laborious, a large fraction of cases turn out to be normal indicating that much of this time is spent diagnosing completely normal cases. This paper describes our early experiences working with Aravind Eye Hospitals to develop an automated system to detect diabetic retinopathy from retinal images. The automated diabetic retinopathy problem is a hard computer vision problem whose goal is to detect features of retinopathy, such as hemorrhages and exudates, in retinal color fundus images. We describe our initial efforts towards building such a system using a range of computer vision techniques and discuss the potential impact on early detection of diabetic retinopathy.

Road traffic accidents (RTAs) are a major public health concern, resulting in an estimated 1.2 million deaths and 50 million injuries worldwide each year. In the developing world, RTAs are among the leading cause of death and injury; Ethiopia in particular experiences the highest rate of such accidents. Thus, methods to reduce accident severity are of great interest to traffic agencies and the public at large. In this work, we applied data mining technologies to link recorded road characteristics to accident severity in Ethiopia, and developed a set of rules that could be used by the Ethiopian Traffic Agency to improve safety.

The gender divide in the access to technology in developing economies makes gender characterization and automatic gender identification two of the most critical needs for improving cell phone-based services. Gender identification has been typically solved using voice or image processing.   However, such techniques cannot be applied to cell phone networks mostly due to privacy concerns. In this paper, we present a study aimed at characterizing and automatically identifying the gender of a cell phone user in a developing economy based on behavioral, social and mobility variables. Our contributions are twofold: (1) understanding the role that gender plays on phone usage, and (2) evaluating common machine learning approaches for gender identification. The analysis was carried out using the encrypted CDRs (Call Detail Records) of approximately 10,000 users from a developing economy, whose gender was known a priori. Our results indicate that behavioral and social variables, including the number of input/output calls and the in degree/out degree of the social network, reveal statistically significant differences between male and female callers. Finally, we propose a new gender identification algorithm that can achieve classification rates of up to 80% when the percentage of predicted instances is reduced.

Local information access (LIA) programs tap into existing public health data flows, and present data in simple and useful ways to ground staff. LIAs hold great potential for improving rural health systems in developing regions; benefits include more evidence-based decision making and optimizations at a local scale, as well as improved service delivery and data quality. Our fledgling LIA program in rural Uganda currently provides clinicians with a small set of static data visualizations for discussion. To increase the program’s effectiveness, we want to automatically identify relevant data visualizations. We propose an adaptive tool that learns from local clinicians’ decision-making processes to predict and generate visualizations that show actionable anomalies.

We explore the prospect of inferring the epicenter and influences of seismic activity from changes in background phone communication activities logged at cell towers. In particular, we explore the perturbations in Rwandan call data invoked by an earthquake in February 2008 centered in the Lac Kivu region of the Democratic Republic of the Congo. Beyond the initial seismic event, we investigate the challenge of assessing the distribution of the persistence of needs over geographic regions, using the persistence of call anomalies after the earthquake as a proxy for lasting influences and the potential need for assistance. We also infer uncertainties in the inferences and consider the prospect of identifying the value of surveying the areas so that surveillance resources can be best triaged.