Data Community DC and District Data Labs are hosting a Supervised Machine Learning with R workshop on Saturday April 30th. Come out and learn about R's capabilities for regression and classification, how to perform inference with these models, and how to use out-of-sample evaluation methods for your models!
Data Community DC and District Data Labs are hosting another session of their Building Data Apps with Python workshop on Saturday February 6th from 9am - 5pm. If you're interested in learning about the data science pipeline and how to build and end-to-end data product using Python, you won't want to miss it. Register before January 23rd for an early bird discount!
Data Community DC and District Data Labs are hosting a Fast Data Applications with Spark & Python workshop on Saturday January 23rd from 9am - 5pm. Register before January 9th for an early bird discount!
For several years now, the DC Nightowls meetup has been a stable of after hours coworking for entrepreneurs, startups, and self-starters doing interesting projects in the Metropolitan DC area. A number of our Data Science community members have been owls as well.
Recently, we decided to combine forces by re-focusing DC Nightowls through a new program called DC2 Digital Nomads. The new program's focus is on the gig economy, freelance knowledge workers, and remote working.
If you haven’t come around yet, it’s past time: Data ethics is really important.
A quick glance at recent ethical dilemmas is telling. Troubling instances of the mosaic effect — in which different anonymized datasets are combined to reveal unintended details — include the tracking of celebrity cab trips and the identification of Netflix user profiles. It is also difficult to remain unconcerned with the tremendous influence wielded by corporations and their massive data stores, most notoriously embodied by Facebook’s secret psychological experiments. And new issues are emerging all the time. I dare you to read MIT’s recent article on why we must train self-driving cars to kill without letting out a disquieted “Huh.”
During an upcoming free workshop, Andrej Lapajne will be going in depth on the benefits of using IBCS to improve your data visualization practices and communication. Here is a brief introduction to what IBCS is and how it is helping businesses across the world visualize their data effectively and consistently.
Are you using data visualization to improve your reports, presentations and communications, or to unknowingly hinder them? All too often, reports fall somewhere between messy spreadsheets and dashboards, full of poorly labeled and inappropriate charts, that simply do not get the message across to the decision-makers.
Countless reports and presentations are created throughout organizations on a daily basis, all in different formats, lengths, shapes and colors, depending on preferences of the person who prepares them. The end results are often managers not making their way through the data presented, time being wasted, and important decisions failing to be made.
The solution - International Business Communication Standards
In 2004 Dr. Rolf Hichert, the renowned German professor, took on a challenge to standardize the way data visualizers present data in their reports, dashboards and presentations. His extremely successful work culminated in 2013 with the public release of the International Business Communication Standards (IBCS) the world’s first practical proposal for the standardized design of business communication.
The IBCS consistently define shapes and colors of actuals and budgets, variances, different KPIs, etc. Often referred to as the “traffic signs for management”, the IBCS are a set of best practices that went viral in Europe and have solved business communication problems in numerous companies such as SAP, Bayer, Lufthansa, Philips, Coca-Cola Bottlers, Swiss Post, etc.
Profit & Loss analysis (income statement) with waterfall charts and variances
How does it work?
Let’s take a look at a typical column chart, designed to help us compare actual sales figures vs. budget:
Is it efficient? The colors used are completely arbitrary, probably just an accidental default of the software tool. It is quite hard to estimate the variances to budget. Are we above the budget or below the budget in a particular month? For how much?
Now let’s observe the same dataset, designed according to the IBCS:
The actuals are depicted as dark grey full columns, while the budget is an outline. This is called scenario coding: the budget is an empty frame that has to be filled up with the actuals.
The variances are explicitly calculated and visualized. Positive variance is green, negative is red. The user’s attention is guided to the variances, which are in this case the key element to understand the sales performance.
The values are explicitly labeled at the most appropriate position on the chart. All texts are standardized, exact, short and displayed horizontally.
Storyline, visual design and uniform notation
The IBCS standards are not just about charts. They comprise of an extensive set of rules and recommendations for the design of business communication that help:
- Organize and structure your content by using an appropriate storyline
- Present your content by using an appropriate visual design and
- Standardize the content by using a consistent, uniform notation.
After you apply the IBCS rules to your standard variance report, it will look something like this:
Sales variance report - Actual vs PY vs Budget
As you may have noticed, this report has several distinctive features:
- The key message (headline) at the top
- Title elements below the key message
- Clear structure of columns (first PY for previous year values, then AC for actual and at the end BU for budget; always in this order)
- Scenario markers below column headers (grey for PY, black for AC and outline for BU)
- Strictly no decorative elements, only a few horizontal lines
- Variances are visualized with red/green “plus-minus” charts and embedded into the table
- Absolute variances (ΔPY, ΔBU) are visualized as a bar chart, while relative variances (ΔPY%, ΔBU%) are visualized as “pin” charts (we prefer to call them “lollipop” charts)
- Semantic axis in charts: grey axis for variance to PY (grey = previous year), double line for variance to budget (outline = budget)
- Numbered explanatory comments that are integrated into the report.
A clear message, appropriate data visualization and accurate explanations. The story that numbers are telling presented on one single page. That's what the managers expect.
We will be going into much further detail on IBCS guidelines for visualizing data and business communications during a free workshop on Oct 29th at 10am. To learn more and register, please visit us at zebra.bi/dc2015.
Data Community DC and District Data Labs are hosting a Data Mining & Machine Learning with R workshop on Saturday October 31st from 9am - 5pm. Register before October 17th for an early bird discount!
We had a packed house on Saturday, Oct 10th for our State of Data Science Education event. The panelists were fantastic and the questions from the crowd were amazing. The tweets below represent just a fraction of the awesomeness that took place. Check out the twitter hashtag #de15 for the whole story.
Data Community DC and District Data Labs are hosting an Intro to Python for Data Science workshop on Saturday October 3rd from 9am - 5pm. Register before September 19th for an early bird discount!
Data Community DC and District Data Labs are excited to be hosting an Introduction to R Programming workshop on Saturday, September 26th. Register before 9/12/2015 for an early bird discount!
Data Community DC is pleased to announce our State of Data Science Education event.
Its goal is to bring together educators, vocational programs and companies to discuss the state and future of data science.
With the rise of data science both schools and companies are rapidly building up their programs and departments, but their beliefs about what a data scientist is and what skills they should have vary considerably.
Last week, the Urban Institute hosted a discussion on the evolving landscape of data and the potential impact on social science. “Machine Learning in a Data-Driven World” covered a wide range of important issues around data science in academic research and their real policy applications. Above all else, one critical narrative emerged:
Data are changing, and we can use these data for social good, but only if we are willing to adapt to new tools and emerging methods.
Last month, a new meetup group for women data scientists in the DC area was started by Mandi Traud and Jackie Kazil.
Women Data Scientists DC is a meetup group for women data scientists, women who want to be data scientists, and supporters of women in data science. Their monthly meetings will include presentations by data scientists, networking events, mentoring opportunities, and workshops to learn new data science skills.
Co-founders Jackie Kazil and Mandi Traud launched on July 9th with two members, and by the next day, the group had more than 85 members and growing!
Here's what the co-founders said individually when asked about how and why they decided to start this group.
This guest blog post on ROCs was spurred by a conversation in the Q&A at Data Science DC’s June 16th Meetup on “Predicting Topics and Sharing in Social Media”. John Kaufhold, managing partner of Deep Learning Analytics, asked Bill Rand, Assistant Professor of Marketing at University of Maryland, about ROCs and convex hulls. In the post, Dr. Kaufhold satirizes data science moments lost in Q&A, talks ROC curves, and discusses the value of error bars in visualizing data science results.