Past Seminar Series

Date: 2017-09-29

Time: 14:30-16:30

Location: BURN 1205

Abstract:

McNeil: In this talk we study a class of backtests for forecast distributions in which the test statistic is a spectral transformation that weights exceedance events by a function of the modelled probability level. The choice of the kernel function makes explicit the user’s priorities for model performance. The class of spectral backtests includes tests of unconditional coverage and tests of conditional coverage. We show how the class embeds a wide variety of backtests in the existing literature, and propose novel variants as well. We assess the size and power of the backtests in realistic sample sizes, and in particular demonstrate the tradeoff between power and specificity in validating quantile forecasts.

Date: 2017-09-22

Time: 14:00-15:00

Location: BRONF179

Abstract:

We study the problem of nonparametric dependence detection. Many existing methods suffer severe power loss due to non-uniform consistency, which we illustrate with a paradox. To avoid such power loss, we approach the nonparametric test of independence through the new framework of binary expansion statistics (BEStat) and binary expansion testing (BET), which examine dependence through a filtration induced by marginal binary expansions. Through a novel decomposition of the likelihood of contingency tables whose sizes are powers of 2, we show that the interactions of binary variables in the filtration are complete sufficient statistics for dependence. These interactions are also pairwise independent under the null. By utilizing these interactions, the BET avoids the problem of non-uniform consistency and improves upon a wide class of commonly used methods (a) by achieving the optimal rate in sample complexity and (b) by providing clear interpretations of global and local relationships upon rejection of independence. The binary expansion approach also connects the test statistics with the current computing system to allow efficient bitwise implementation. We illustrate the BET by a study of the distribution of stars in the night sky and by an exploratory data analysis of the TCGA breast cancer data.

Date: 2017-09-15

Time: 15:30-16:30

Location: BURN 1205

Abstract:

The demand for time series forecasting at Google has grown rapidly along with the company since its founding. Initially, the various business and engineering needs led to a multitude of forecasting approaches, most reliant on direct analyst support. The volume and variety of the approaches, and in some cases their inconsistency, called out for an attempt to unify, automate, and extend forecasting methods, and to distribute the results via tools that could be deployed reliably across the company. That is, for an attempt to develop methods and tools that would facilitate accurate large-scale time series forecasting at Google. We were part of a team of data scientists in Search Infrastructure at Google that took on the task of developing robust and automatic large-scale time series forecasting for our organization. In this talk, we recount how we approached the task, describing initial stakeholder needs, the business and engineering contexts in which the challenge arose, and theoretical and pragmatic choices we made to implement our solution. We describe our general forecasting framework, offer details on various tractable subproblems into which we decomposed our overall forecasting task, and provide an example of our forecasting routine applied to publicly available Turkish Electricity data.

Date: 2017-09-08

Time: 15:30-16:30

Location: BURN 1205

Abstract:

A central goal of population genetics is the inference of the biological, evolutionary and demographic forces that shaped human diversity. Large-scale sequencing experiments provide fantastic opportunities to learn about human history and biology if we can overcome computational and statistical challenges. I will discuss how simple mid-century statistical approaches, such as the jackknife and Kolmogorov equations, can be combined in unexpected ways to solve partial differential equations, optimize genomic study design, and learn about the spread of modern humans since our common African origins.

Date: 2017-04-06

Time: 15:30-16:30

Location: Universite Laval, Pavillon Vachon, Salle 3840

Abstract:

Instrumental variable (IV) methods are popular in non-experimental studies to estimate the causal effects of medical interventions or exposures. These approaches allow for the consistent estimation of such effects even if important confounding factors are unobserved. Despite the increasing use of these methods, there have been few extensions of IV methods to censored data regression problems. We discuss challenges in applying IV structural equational modelling techniques to the proportional hazards model and suggest alternative modelling frameworks. We demonstrate the utility of the accelerated lifetime and additive hazards models for IV analyses with censored data. Assuming linear structural equation models for either the event time or the hazard function, we proposed closed-form, two-stage estimators for the causal effect in the structural models for the failure time outcomes. The asymptotic properties of the estimators are derived and the resulting inferences are shown to perform well in simulation studies and in an application to a data set on the effectiveness of a novel chemotherapeutic agent for colon cancer.

Date: 2017-03-31

Time: 15:30-16:30

Location: BURN 1205

Abstract:

In modern scientific research, massive datasets with huge numbers of observations are frequently encountered. To facilitate the computational process, a divide-and-conquer scheme is often used for the analysis of big data. In such a strategy, a full dataset is first split into several manageable segments; the final output is then aggregated from the individual outputs of the segments. Despite its popularity in practice, it remains largely unknown that whether such a distributive strategy provides valid theoretical inferences to the original data; if so, how efficient does it work? In this talk, I address these fundamental issues for the non-parametric distributed kernel regression, where accurate prediction is the main learning task. I will begin with the naive simple averaging algorithm and then talk about an improved approach via ADMM. The promising preference of these methods is supported by both simulation and real data examples.

Date: 2017-03-24

Time: 15:30-16:30

Location: BURN 1205

Abstract:

Sample size determination problem is an important task in the planning of clinical trials. The problem may be formulated formally in statistical terms. The most frequently used methods are based on the required size, and power of the trial for a specified treatment effect. In contrast to the Bayesian decision-theoretic approach, there is no explicit balancing of the cost of a possible increase in the size of the trial against the benefit of the more accurate information which it would give. In this talk a fully Bayesian approach to the sample size determination problem is discussed. This approach treats the problem as a decision problem and employs a utility function to find the optimal sample size of a trial. Furthermore, we assume that a regulatory authority, which is deciding on whether or not to grant a licence to a new treatment, uses a frequentist approach. The optimal sample size for the trial is then found by maximising the expected net benefit, which is the expected benefit of subsequent use of the new treatment minus the cost of the trial.

Date: 2017-03-17

Time: 15:30-16:30

Location: BURN 1205

Abstract:

We consider the asymptotic consistency of maximum likelihood parameter estimation for dynamical systems observed with noise. Under suitable conditions on the dynamical systems and the observations, we show that maximum likelihood parameter estimation is consistent. Furthermore, we show how some well-studied properties of dynamical systems imply the general statistical properties related to maximum likelihood estimation. Finally, we exhibit classical families of dynamical systems for which maximum likelihood estimation is consistent. Examples include shifts of finite type with Gibbs measures and Axiom A attractors with SRB measures. We also relate Bayesian inference to the thermodynamic formalism in tracking dynamical systems.

Date: 2017-03-10

Time: 15:30-16:30

Location: BURN 1205

Abstract:

The immune system does a lot more than killing “foreign” invaders. It’s a powerful sensory system that can detect stress levels, infections, wounds, and even cancer tumors. However, due to the complex interplay between different cell types and signaling pathways, the amount of data produced to characterize all different aspects of the immune system (tens of thousands of genes measured and hundreds of millions of cells, just from a single patient) completely overwhelms existing bioinformatics tools. My laboratory specializes in the development of machine learning techniques that address the unique challenges of high-throughput single-cell immunology. Sharing our lab space with a clinical and an immunological research laboratory, my students and fellows are directly exposed to the real-world challenges and opportunities of bringing machine learning and immunology to the (literal) bedside.

Date: 2017-02-24

Time: 15:30-16:30

Location: BURN 1205

Abstract:

This talk will provide an introduction to the first of the pillars in Stephen Stigler’s 2016 book The Seven Pillars of Statistical Wisdom, namely “Aggregation.” It will focus on early instances of the sample mean in scientific work, on the early error distributions, and on how their “centres” were fitted.

Speaker

James A. Hanley is a Professor in the Department of Epidemiology, Biostatistics and Occupational Health, at McGill University.