Our actions, movements and decisions are all based on the summoning of past experiences from our memory. We then evaluate and integrate these experiences within a given context to estimate the likelihood of a set of outcomes that might result from any particular action. Essentially, people live their lives predicting the results of actions based on the evaluation of past experiences. These experiences are the “data” for the decision-making processes, and the evaluation of these experiences and outcomes is the “methodology” of our own individual “predictive modeling” process. For example, experience has taught us that it is very likely that if we touch a hot frying pan we shall burn our fingers.

Most events in life, have variable probabilities of occurring depending on the particular set of circumstances existent at the time of their happening. And our ability to predict what might happen in any particular situation is a function of our own experiential knowledge base in addition to these external factors. Thus, the limitations of our own experiences coupled with the lack of complete correspondence among events and outcomes limit the ability to predict what might happen in any given circumstance. What, then, can we do to reduce the uncertainty of future events?

Predictive Modeling and Parking Facility Security

Security decision-making that makes business sense in the parking facility environment, or in any other security-relevant environment, requires the objective, valid and reliable assessment of the risks of property loss and personal injury in various scenarios and circumstances at any given site. This assessment should not be based on the personal experiences of the security manager, which is a subjective impression. Analysis of valid and accurately measured, reliable data will determine the best action to take. Similar kinds of incidents yield the same kinds information each time about security loss incidents and the neighborhood characteristics that produce crime risks at and around that site.

Incident-based and crime risk data then need to be related to the security measures that were in place at the site of the event. Then, these relationships between incident and security measures should be analyzed in a multivariate way, using statistical modeling techniques to determine how the property loss and personal injuries are affected by the security measures utilized. The replication of this process across sites and time produces the basis for the development of a predictive or “forecasting” model, upon which security resource allocation can be rationally implemented in a cost-effective, valid manner.

The Predictive Model

A predictive model in the parking facility security context is an equation or combination of equations which relate a set variables (predictor) to an outcome or set of outcomes with the probability of those outcomes specified by the model according to how tight the fit is among predictor and outcome variables in the collected data. These methods are well documented in the literature of statistical methods and are found under the general heading of “regression-based” techniques.  

SIDEBAR: Benefits of Predictive Modeling

Many business benefits can be realized through the use of these models for parking facility resource management:

• Resource allocations decisions are based on valid and reliable data rather than upon subjectively based measures that depend solely upon the experiences of individuals (which are biased on a person-by-person basis).
  
  
• Property and personal injury losses can be predicted with a statistically valid confidence bound.
  
  
• The costs of security implementations can be balanced against the likelihoods of projected losses.
  
  
• As a result, cost benefit decisions can be utilized in the security function of a business.
  
  
• Contemporary business managers need data based budgets for fiscal planning from all divisions of the company, including security.
  
  
• The effectiveness of security measures in reducing losses can be dramatically improved through the use of empirically based objective, evaluative techniques, which help determine security activities that reduce losses.