Why are so many business segments utilizing Data Science in their business decisions? These business segments include but are not limited to marketing, transportation, telephony, medicine, insurance, finance, entertainment and many others.
Editor’s Note: Bruce Hahn, CCIM, CRE, MAI, SRA is kind enough to write another Guest Post for us this week. To find out more about Bruce Hahn check out his website at BruceHahn.Com.
What is it about Data Science that is so useful to these businesses?
Data Science can:
Reduce inefficiencies
Predict Trends and Consumer Behavior
Develop Market Understanding
Improve Hiring Practices
Aid in Research About Competitors
Test Business Initiatives
Increase Security
Make Predictions for Business Outcomes
Data Science is an interdisciplinary field that includes a mix of computer processing power and programming, statistics, and mathematics with an overlapping domain knowledge (or specific expertise). Data science allows us to understand what happens, why it happens, and enables us to make informed forecasts about what is likely to happen in the future. Data science enables users to strategize and make decisions based on data that exists, but might otherwise be overlooked or underutilized.
Data Science utilizes analytics to understand all the data that is available. This is frequently based on simple descriptive statistics that explain a population of data – not just a sample of data like inferential statistics. This is the easier to understand branch of statistics that we learned in high school. This includes concepts that are intuitive like distribution, variance, correlation, and measures of central tendency. Powerful data visualizations aid in understanding these details, but these illustrations are not readily available outside Data Science. Data Science also allows Diagnostic Analysis and Predictive Analytics.
Diagnostic analysis tools can be easy to understand in concept, but data science uses large amounts of computing power to crunch large amounts of data to obtain meaningful results that would be difficult to process using manual computation. Powerful data visualizations make it easy to understand this analysis. Many of these tools are simple and easy to understand – like linear regression. Bayesian probability analysis is not that complicated a concept, but it requires a significant amount of mathematical processing that is not practical without computer processing power. Even machine or statistical learning is easy enough to understand – simple models use measures of closest distance from other data points. Again, an easy to understand concept but it too needs significant amounts of computer processing.
The results of diagnostic analysis in Data Science can lead an analyst to develop predictive models to show informed decision making about specific future outcomes and their likeliness. The significance of Data Science to so many business segments is its power to provide insights into consumer behavior, market behavior, and to make meaningful forecasts about future trends from data analytics. These are the reasons so many business segments are already utilizing Data Science. It is exactly what real estate appraisers set out to do every day – to analyze the behavior of real estate buyers and sellers, the real estate market and to make forecasts about real estate values. So why aren’t appraisers utilizing Data Science in their professional work? More importantly, why aren’t Data Science principles included in modern Appraisal Education and Methodology?
October 12, 2022 @ 10:03 am
Good article.
October 17, 2022 @ 1:25 pm
Nice Article. Data analysis is a tool for appraiser’s it is not a tool that should be used to replace appraisers. Why? Please watch the Nova series 10/12/2022. Computers VS Crime. What this program found, is that data science is inherently flawed. Has a built-in Bias. Experts state that in predictive analysis that it’s impossible for a mathematical algorithm to be UNBIASED. A random group of data is just as reliable as an Algorithm. This data science is built off linear regression. Hiring practices, look at what Amazon found in their Algorithms which were based on AI (machine learning). Inherent bias against Women. Why use it. It’s cheap. Wrongly believed to be unbiased. Data is incorrectly used by the participants who think they can’t be sued, secret algorithm.
October 17, 2022 @ 3:05 pm
Nice article. It’s always about money (cheap data) or otherwise, why would the use of data science be mis-used so much. Data science should be in the hands of the correct expert, in this case it would be a Real Estate Appraiser. Decisions on value should be left up to the appraiser. Not the computer or analyst. Please view the program on Nova. Called Computer VS Crime. In our case, computers vs appraisers. Dated 10/12/2022. In this show, Mathematicians and other experts show how the algorithms are inherently BIASED. The fallacy of big data as unbiased. In fact, as noted in the program. Amazon hired the best data people to come up with an algorithm for hiring people (as mentioned above). Over and over their algorithm showed a BIAS against WOMEN. They even used AI (artificial intelligence) with 50,000 data points! What is the reliability? Nova shows the reliability is no better than a random group of people. Mathematicians say it’s impossible for an algorithm to make a decision that is UNBIASED on predictive models. So, once again, this is a TOOL for appraisers. Not a TOOL to replace appraisers. This is a legal mine field.
June 24, 2023 @ 7:30 pm
Why are appraisers not using data science? 1) Good data in abundance is not available to appraisers. Our data is sparse, and very noisy. An appraiser can’t afford to review 40 sales and check for non-arm’s length sales, mis-classified sales, data errors, etc. 2) The meaning of some variables is difficult to quantify, e.g. location. Distance from downtown, or from an arterial crossroads? Traffic count? Is is better to be on an arterial, or to be tucked away in a nice business park? 3) Very few appraisers have the skills for data science. If they did, they’d be making more money as a data scientist. 4) Appraisal is formalistic and regulated. You can do all the data science you want, but clients and regulators still expect 4-5 comparable sales and an adjustment grid of made-up adjustment factors. Which brings us to 5) Price. You can spend many hours on statistical analysis of numerous sales, but in the end, you’ll still only get $2500 – $3000 for a commercial report (or $500 for SFR). That’s how it is. Data science will be pursued in the appraisal waiver/AVM world. These products will be alternatives to appraisals, not appraisals, and will be prepared by data companies or the GSEs, not appraisers.