is a specific case of predictive analytics, but there are many other conditions, patterns, or behaviors that you may want to predict other than understanding how trends will continue into the future.

This is a specific type of regression where the dependent variable is binary (zero or one), and the output of the regression model is the probability of an event occurring (a value between zero and one).

This category of predictive models is used to find groups of observations in the data that behave similarly, or share similar characteristics, and then categorize future observations accordingly.

is the most common use of cluster analysis in hospitality.

This term refers to a large class of models that are specifically designed to find patterns in large data sets when the researcher does not have a predetermined idea of potential relationships in the data.

a broader category of modeling similar to clustering, where the - algorithm evaluates observations to determine what group they belong to.

a form of predictive modeling, where a model is created to predict the value of a target variable based on several input variables. It overcome the limitations of some predictive models, which can consist of the number or type of variables that can be included, making them very useful for extremely large data sets.

determines the users' intention from their behavior in interacting with a computer system, particularly a search engine.

determines the users' intention from their behavior in interacting with a computer system, particularly a search engine.

is a specific mathematical technique that solves for the decisions you need to make to get to the best possible answer to a specific objective, accounting for all of the constraints unique to the problem.

the equation that describes the desired outcome of the problem, always written as a goal of minimizing or maximizing.

these are the outputs of the optimization problem. They represent the decisions the model recommends to achieve the best possible outcome.

are how you express the operating conditions under which the problem should be solved.

is a term you have likely heard associated with a wide variety of problems. and has probably been pitched by some of the analytics vendors you have come in contact with.

can be used to mine the content of any unstructured text document, either created on external sites like Facebook, Twitter, or TripAdvisor, or created internally like call logs or open-ended questions on a guest survey.

is most effective when it is applied natively, as opposed to translated text, which is why it is important for a global industry like hospitality to work with software that has the largest available portfolio of languages in their text analytics package.

this identifies key topics and phrases in electronic text and sorts them into categories. It eliminates the manual work of reading and tagging documents, giving you much faster results. Text documents can be organized and tagged for search, making it easier to find, sort, or process the content.

it uncovers related concepts in large volumes of conversations. It surfaces key topics that can be used in future analyses, like predicting or understanding guest behavior. mo

this helps you understand guest opinions by applying natural language processing to the text documents. It identifies how guests feel -positive, negative, or neutral-about key attributes of your product, brand, or service

calculations are split across multiple central processing units (CPUs) to solve a bunch of smaller problems in parallel, as opposed to one big problem in sequence.

most analytics programs lift data sets out of the database, execute the "math," and then dump the data sets back in the database.

this capability is a bit harder to understand for nontechnical people, but it provides a crucial advantage for both reporting and analytics. Large sets of data are typically stored on the hard drive of a computer, which is the physical disk inside the computer (or server). It takes time to read the data off the physical disk space, and every pass through the data adds additional time (note that the math behind the analytics requires multiple passes through data).

these cloud-based solutions are complete end-to-end software packages that are standardized for the entire user base. While they can be configured, generally, the software is the same for all companies that access it

solution is a proprietary solution for each customer that is effectively the same as having an on-premise solution, except it is located in a remote data center and managed by the vendor.

the vendor assumes all responsibility for maintaining the environment, including regular maintenance, upgrades, data transfers, or managing memory and storage capacity.

cloud providers guarantee a minimum up-time for the solution, and build in the mechanisms required to achieve this. You have the security of knowing the system will be available when you need it.

particularly for SaaS applications, the time to implementation is dramatically reduced for cloud solutions.

most cloud providers have extensive data encryption and cyber security measures in place. In fact, some financial services companies use hosted solutions to manage credit card fraud activity, and are able to meet all of their both internal and external regulatory security measures to ensure that credit card numbers are not stolen, and the consumer's privacy is respected.

an on-premise solution is 100% in the control of the company that uses it. The data flow in and out is controlled by the company, as is the system configuration. Many companies see their data as a strategic asset, and do not want it to be managed by a third party.

some industries' data is highly regulated (casinos, for example). and these regulations prohibit some data from ever leaving the facility where it is created. While the data centers for cloud solutions may have every protection mechanism in place, these companies are simply not allowed to leverage them. This is particularly an issue for customer data.

while the up-front cost of many on-premise solutions may be higher, the long- term costs can be significantly less. For hosted solutions, you can pay all of the up-front hardware and software costs plus an ongoing maintenance fee for the solution. than up- front software costs for an on-premise solution.

The move to digital has given hospitality companies a platform to interact with guests in real time, or near real time, as guests are browsing the web or using an app

is now being used across hospitality and travel. Many interactions are based on business rules, such as "if the guest clicks here, show them this banner ad.

typically work by understanding the context of the interaction, and looking up a score or analytic result that is matched to an offer or communication sent back to the guest in the moment.

in hospitality and gaming represent a wide variety of guest facing functions on property, as well as the supporting infrastructure at the regional or corporate level. Such as the front desk, concierge, and housekeeping, as well as food and beverage, banquets and catering, retail, spa, and other related services and, pricing and revenue management for all outlets

They capture labor costs for the line-level employees, and the costs associated with ordering the various products used in the delivery of service

in this context means the number of people served or number of "jobs" executed at any given service area.

are generally very well aware of all of the factors that might influence demand for all of the outlets across the enterprise.

refers both to the time of day that something happens

refers to the amount of time it takes to complete a service

tells us that it is the variability in the service time-these occasions when the service deviates from typical patterns-that creates lines. disgruntled customers, or idle employees eating up payroll.

have evolved over time, but the value of understanding reactions to the service experience remains the same.

is probably the most useful, and widely used, advanced analytic application for operations.

this forecasting method predicts the amount of demand that existed for your service, regardless of whether there was space to accommodate it.

by this capacity - in the historical data, you will never actually "see" more demand than was able to be served, even if the restaurant turned guests away.

is a useful forecasting method in the cases where there are small amounts of data that roll up to an overall number.

This method can help balance the need for more detailed forecasts with the challenge of decreasing accuracy as the number of observations becomes smaller.

allows you to model complex real world systems in an artificial environment so you can test options before you implement them and they impact the guest experience.

One of the more interesting predictive analytic applications that is not yet widely used in hospitality or gaming operations.

is a branch of analytics that predicts the behavior of lines. As it turns out, this modeling is among the most complex mathematical techniques, because it needs to account for uncertainty.

can predict what will happen under a variety of service conditions, ultimately determining the best way to design a service process to better manage waits

this is the group from which guests are obtained. It isn't homogeneous. It is comprised of many market segments and many guests or patrons with different needs.

this is the rate at which guests arrive to the service system and if necessary join the queue. This can vary by time period.

this refers to the way the queue is designed. For example, single line, single server, a snake line to multiple servers, and multiple lines to multiple servers.

this refers to the rules that govern the order in which guests or patrons are selected from the line. For example, the available server could simply take the next guest in line, or pass over that next guest in favor of a guest with a more immediate need or a higher status.

this final feature describes the process of delivering service. This could range from self-service to a multistage server process.

what drives the amount of work-is it cars to park, rooms to clean, and guests to check

for each entity, does the time it takes to serve vary significantly enough that the duration and the variability around it should be accounted for?

how much does demand vary by time period and how do you - schedule your labor? Can you stagger shift start and end times? Do you have part-time workers who can cover busy periods?

operations departments that are in early stages of analytical capability operate in a highly manual environment.

the typical operations department is able to look at their data at a pretty detailed level.

the most sophisticated operations groups in hotels and casinos are heavy users of very sophisticated forecasting techniques.