Topics in Computing for Data Science on STATS/BIODS 352

2022 Schedule

Mon, 01 Jan 0001 00:00:00 +0000

Some minor changes/updates are possible and so we advise you to check this page often.

We will use Canvas for class announcements, materials and other administrivia.

Most meetings are in person, although a few will be remote as will be noted in due course. The class meets Thursdays, 11:00-12:00 noon, in 200-034 (Lane History Corner). Search for 200-034 on Campus Map. (As you enter the History Corner from the main quad, the room is on the ground floor, down the stairs and to the far end, beyond room 30 and is labelled 34!)

Instructors

Mon, 01 Jan 0001 00:00:00 +0000

John M. Chambers, Department of Statistics

Balasubramanian Narasimhan, Department of Statistics and Department of Biomedical Data Sciences

Motivation

Mon, 01 Jan 0001 00:00:00 +0000

Modern data science finds use in wide-ranging applications in many fields. Data science training involves sufficient grounding in the core components—Applied Mathematics, Statistics and Computer Science. The practice of data science invariably requires sufficient mastery of computing tools in pursuit of research goals.

Increasingly, methodological advances come with computing tools/software implementations that are not always simple or even familiar to the all data scientists. That is often due to the nature of the beast: varied applicability, software choice, and sheer complexity. Examples include techniques for privacy-aware computations, debiasing methods for algorithms, robust inference, causal modeling, to name just a few. Some methods are also bespoke and require practical experience to discern the full applicability and trade-offs involved in utilizing them for research. So there are many data scientists who are aware of such techniques but have never dipped their toes in the water.

Participation

Mon, 01 Jan 0001 00:00:00 +0000

Where appropriate, this seminar course will progress in groups of two consecutive meetings on a specified topic relevant to computing for data science. In the first meeting, the speaker will introduce the topic and work through a simple example. The participants will be expected to replicate the example outside of class before the next meeting (see below for details). The second meeting will involve a discussion of the topic, including any issues that arose with the example. The speaker may use the last half of the meeting to suggest further applications and reading/tutorial material.

Schedule

Mon, 01 Jan 0001 00:00:00 +0000

Any changes to the schedule will be reflected here, so we advise you to check this page often.

We will use Canvas for class announcements, materials and other administrivia.

The class meets Thursdays, 10:30-11:50, in Green Earth Sciences, Room 150

April 6, 13



Ryan Tibshirani	Daniel J. McDonald
UC Berkeley	U of British Columbia

Opportunities and Challenges in Auxiliary Surveillance for Public Health the United States

In 2015, the Delphi group at Carnegie Mellon University launched an effort called the Epidata project, to collect and make publicly available signals that reflect infectious disease activity in real-time. The focus was primarily on seasonal influenza in the United States. In March 2020, this effort was massively expanded and accelerated to help support the COVID-19 response. Now, Epidata has over 4.5 billion records, with ~3 million records added daily, and receives between 100,000 and 1 million API queries per day. It covers a diverse set of data streams, both novel and traditional, for tracking COVID-19, influenza, and other diseases. The first lecture, on April 6, will give a high-level summary of the main goals behind Epidata, and the challenges and opportunities in auxiliary surveillance for public health. The second lecture, on April 13, will dive into some of the software packages that Delphi is building that support data access, as well as nowcasting and forecasting.

Schedule

Mon, 01 Jan 0001 00:00:00 +0000

Any changes to the schedule will be reflected here, so we advise you to check this page often.

We will use Canvas for class announcements, materials and other administrivia.

The class meets Wednesdays, 1:30–2:50pm, in Hewlett Teaching Center, Room 101. (The location was changed after the first class.)

April 3, 24

Balasubramanian Narasimhan (Stanford University)

Containers, Workflows and Tools for HPC

Efficient High Performance Computing demands robust workflows and tools that let scientists “do the right thing” as easily as possible. Those right things include removal of drugery by recognizing repeated patterns to be exploited, yet allowing for the inevitable changes, while paying close attention to issues of reproducibility. I will discuss a number of tools that make this possible and also delve into virtualization using containers, which are essentially virtual machines or collections of them that can moved to on-prem or cloud infrastructures. These techniques will find use both in the existing Stanford HPC infrastructure (including the soon-to-be GPU cluster) and elsewhere. No background will be assumed, and I will start from the basics. These lectures will be very hands-on and details on open-source software tools that need to be installed will be provided in due course.

Schedule

Mon, 01 Jan 0001 00:00:00 +0000

Any changes to the schedule will be reflected here, so we advise you to check this page often.

We will use Canvas for class announcements, materials and other administrivia.

The class meets Wednesdays, 1:30–2:50pm. The location STLC 105.

April 2

Balasubramanian Narasimhan (Stanford University)

Marlowe: Stanford’s new GPU Cluster

We introduce Stanford’s new GPU-based computational instrument, Marlowe. We will discuss Marlowe’s role in high performance computing infrastructure on campus, what makes it different from existing clusters, and policies that govern its use,. Finally we will delve into software and tools available on the system, and focus on best practice, workflows, all illustrated with examples.

Topics

Mon, 01 Jan 0001 00:00:00 +0000

A variety of computations are essential for successful data science, including but not limited to traditional statistical and numerical methods. This course will examine current capabilities and challenges in data science including techniques important for: obtaining and organizing data; discovering and examining structure; valid explicit inference and prediction; and communicating scientific information effectively, both for policy advice and for general understanding.

The material may involve a variety of programming languages (e.g., R and Python), development environments (e.g., RStudio, Jupyter) and computing facilities (e.g., Cloud, GPU). Individuals are unlikely to be expert over this wide range, but our approach emphasizes being open to the best computing for the task and using modern tools to integrate facilities for a research goal.