April 18, 2021

How Can Financial Institutions Prepare for AI Risks?

In exploring the potential risks of AI, the paper provided “a standardized practical categorization” of risks related to data, AI and machine learning attacks, testing, trust, and compliance. Robust governance frameworks must focus on definitions, inventory, policies and standards, and controls, the authors noted. Those governance approaches must also address the potential for AI to present privacy issues and potentially discriminatory or unfair outcomes “if not implemented with appropriate care.” In designing their AI governance mechanisms, financial institutions must begin by identifying the settings where AI cannot replace humans. “Unlike humans, AI systems lack the judgment and context for many of the environments in which they are deployed,” the paper stated. “In most cases, it is not possible to train the AI system on all possible scenarios and data.” Hurdles such as the “lack of context, judgment, and overall learning limitations” would inform approaches to risk mitigation, the authors added. Poor data quality and the potential for machine learning/AI attacks are other risks financial institutions must factor in.

How to turn everyday stress into ‘optimal stress’

What triggers a stress response in one person may hardly register with another. Some people feel stressed and become aggressive, while others withdraw. Likewise, our methods of recovery are also unique—riding a bike, for instance, versus reading a book. Executives, however, aren’t usually aware of their stress-related patterns and idiosyncrasies and often don’t realize the extent of the stress burden they are already carrying. Leadership stereotypes don’t help with this. It’s no surprise that we can’t articulate how stress affects us when we equate success with pushing boundaries to excess, fighting through problems, and never admitting weakness. Many people we know can speak in detail about a favorite vacation but get tongue-tied when asked what interactions consistently trigger stress for them, or what time of day they feel most energized. To reach optimal stress, we need to be conscious of our stress; in neurological terms, it’s the first step toward lasting behavior change. As the psychiatrist and author Daniel Siegel writes, “Where attention goes, neural firing flows and neural connection grows.”9 And it is these newly grown neurological pathways that define our behavior and result in new habits.

How to Empower Transformation and Create ROI with Intelligent Automation

CIOs see ROI delivered in multiple ways. For example, a recent Forrester study identified that Bizagi’s platform offered 288% financial returns. CIOs seek benefits other than cost savings, such as increased net promoter scores, realized upsell opportunities, and improved end-user productivity gains. ... Only that automation sets a very high bar on what machines can perform reliably, especially when employees often interpret automation to mean “without any human involvement.” For example, you can automate many steps in a loan application and its approval processes when the applicant checks all the right boxes. However, most financial transactions have complex exceptions and actions that require orchestration across multiple systems. Managers and employees know the daily complications and oversimplifying their jobs with only rudimentary automations often leads to a backlash from vocal detractors. That’s why CIOs and IT leaders need more than simple task automation, departmental applications, or one-off data analysis. Digital leaders recognize the importance of intelligence and orchestration to modernize workflows, meet customer expectations, leverage machine learning capabilities, and enable implementing of the required business rules.

Understand Bayes’ Theorem Through Visualization

Before going to any definition, normally Bayes’ Theorem are used when we have a hypothesis and we have observed some evidence and we would like to know the probability of the hypothesis holds given that the said evidence is true. Now it may sound a bit confusing, let’s use the above visualization for a better explanation. In the example, we want to know the probability of selecting the female engineer given who has finished Ph.D. education. The first thing we need is the probability of selecting the female engineer from the population without considering any evidence. The term P(H) is called “prior”. ... As we know Bayes’ theorem is branching from Bayesian statistics, which relies on subjective probabilities and uses Bayes’ theorem to update the knowledge and beliefs regarding the events and quantities of interest based on data. Hence, based on some knowledge, we can draw some initial inferences on the system (“prior” in Bayes) and then “update” these inferences based on the data and new data to obtain the “posterior”. Moreover, there are terms like Bayesian inference and frequentist statistical inference, which is not covered in this article. 

Leveraging Geolocation Data for Machine Learning: Essential Techniques

Fortunately, we don’t have to worry about parsing these different formats and manipulating low-level data structures. We can use the wonderful GeoPandas library in Python that makes all this very easy for us. It is built on top of Pandas, so all of the powerful features of Pandas are already available to you. It works with GeoDataFrames and GeoSeries which are “spatially-aware” versions of Pandas DataFrames and Series objects. It provides a number of additional methods and attributes that can be used to operate on geodata within a DataFrame. A GeoDataFrame is nothing but a regular Pandas DataFrame with an extra ‘geometry’ column for every row that captures the location data. Geopandas can also conveniently load geospatial data from all of these different geo file formats into a GeoDataFrame with a single command. We can perform operations on this GeoDataFrame in the same way regardless of the source format. This abstracts away all of the differences between these formats and their data structures.

Why Probability Theory is Hard

First, probability theorists don’t even agree what probability is or how to think about it. While there is broad consensus about certain classes of problems involving coins, dice, coloured balls in perfectly mixed bags and lottery tickets, as soon as we move into practical probability problems with more vaguely defined spaces of outcome, we are served with an ontological omelette of frequentism, Bayesianism, Kolmogorov axioms, Cox’s theory, subjective, objective, outcome spaces and propositional credences. Even if the probationary probability theorist is eventually indoctrinated (by choice or by accident of course instructor) into one or other school, none of these frameworks is conceptually easy to access. Small wonder that so much probabilistic pedagogy is boiled down to methodological rote learning and rules of thumb. There’s more. Probability theory is often not taught very well. The notation can be confusing; and don’t get me started on measure theory. The good news is that in terms of practical applications, very little can get you a very long way. 

Read more here ...