PASS BA Marathon: September Edition 2017
Self-Service Doesn't Mean You Have to Go It Alone: How to turn analytic silos into analytic teams
Event Date: 27-09-2017 18:00 - Category: Breakout Session (60 minutes) - Track: Analyze
Speaker(s): Julie Hyman
Title: Self-Service Doesn't Mean You Have to Go It Alone: How to turn analytic silos into analytic teams
During this session, product manager Julie Hyman will show you how to create a shared, collaborative environment for your data analytics and data prep teams – turning analytic silos into teams.
She will share how you can easily:
• Document your database connections and key relationships
• Use curated datasets for key information areas
• Share common transformation routines and calculated fields
• Move key workflows off of individual computers and into managed environments
Julie will demonstrate this utilizing Toad Data Point and Toad Intelligence Central from Quest.
Text Analytics Case Study: Bank’s Corporate Responsibility Reports
Event Date: 27-09-2017 19:00 - Category: Breakout Session (60 minutes) - Track: Analyze
Speaker(s): Mark Wilcock
Title: Text Analytics Case Study: Bank’s Corporate Responsibility Reports
The analysis of text documents is rapidly growing in importance and not just of social media but also for legal, academic and financial documents. We'll use a case study based on the analysis of a bank's corporate responsibility reports to learn some techniques; frequency analysis, finding words and phrases specific to one or a few documents in a collection, and key phrase detection. We'll use a variety of tools; R, text analytics web services and Power BI.
Exploring Public Data: An Alternative Data Source
Event Date: 27-09-2017 20:00 - Category: Breakout Session (60 minutes) - Track: Analyze
Speaker(s): Craig Danton
Title: Exploring Public Data: An Alternative Data Source
Public data is a valuable, though underused, alternative data source. From oil and gas production data, to insight into the contents of all shipping containers imported to the United States, public data can yield unexpected market intelligence. Craig Danton, VP of Product at Enigma Technologies, will discuss the extensive volume of public data that financial institutions can leverage to inform their decision-making, as well as the challenges around collecting and harmonizing this data. He will also explain why public data is among the most valuable sources of alternative data, capable of providing complete and from-the-source information to institutions seeking a competitive advantage.
Visualizing, Analyzing & Acting on FinTech Data with GPUs
Event Date: 27-09-2017 21:00 - Category: Breakout Session (60 minutes) - Track: Analyze
Speaker(s): Justin Sears
Title: Visualizing, Analyzing & Acting on FinTech Data with GPUs
From market data to customer data to the back office, financial services firms generate, consume and trade in data in ways that no other industry does. Bill can discuss how the FinTech industry can leverage GPU-analytics to fundamentally transform their clients’ experiences, capitalize on evolving market conditions, prevent fraud, and mitigate risk using the vast amounts of customer, product, and market data at their disposal.
Leveraging Multi-Source & Multi-Type Data to Estimate Organization-Specific Exposure to Executive Risk
Event Date: 27-09-2017 22:00 - Category: Breakout Session (60 minutes) - Track: Analyze
Speaker(s): Dr. Andrew Banasiewicz
Title: Leveraging Multi-Source & Multi-Type Data to Estimate Organization-Specific Exposure to Executive Risk
Given the above, this presentation will outline the custom developed executive risk modeling approach focused on the following three goals:
- To translate an organization’s unique risk profile into an objective, data-derived, multi-attribute-sourced exposure to D&O and other executive risks;
- To uncover the most impactful leading indicators of individual risks, with the goal of contributing to organizations’ risk mitigation efforts;
- To help pinpoint the optimal risk transfer–risk retention structure, taking into account organization-specific financial circumstances, risk appetite and empirically-derived and probability-adjusted net value of coverage.
The analytic process encompasses all publicly-available data, including companies’ own public filings, regulatory actions, claims and subsequent dispositions, notable events (e.g., mergers & acquisitions), press releases and other verbal-to-print announcements, disclosures and other public statements. Method-wise, we make use of:
- Traditional multivariate statistical analyses and machine learning approaches focused of numeric data
- Text mining methodologies focused on the identification of empirically-relevant triggers