Machine Learning - What is it and what is its impact on the Hospitality Industry?

Machine Learning is one of the buzzwords of the moment – ​​but what exactly is it? There are several possible definitions and they vary according to the purpose and associated theme, but, in general, they all converge on the idea that it is one of the methods to achieve Artificial Intelligence (the ability given to machines to imitate intelligent human behavior ). Thus, Machine Learning concerns the development of computational programs that are automatically capable of improving and learning from experience, without being specifically instructed to do so .

However, this is not a concept of the future. It's not limited to cars that don't need drivers, robots that will replace humans or Star Trek-style gadgets . The reality is that we already use Machine Learning in our lives every day and we often don't even notice it . Every time we use Google.com to perform any type of search, Google, using our location, our previous searches and our social media data, applies Machine Learning to give us the most appropriate answer. Our cell phones, with their countless sensors, are capable of predicting what we are doing (whether we are sitting, running, walking, etc.), recording our activities. The same cell phones predict what we are writing to suggest complete words... Among many other daily uses that we make of Machine Learning .

In business applications, Machine Learning is commonly used to build predictive and data analysis models/algorithms, in this case normally referred to as Predictive Analytics .

Machine Learning is typically implemented to address three types of problems:

• Supervised learning : when the application makes use of a set of data with its inputs and already desired outputs (labels), with the aim of learning from examples (e.g. using reservation variables, including which ones were canceled, it is possible to build a model that is capable of predicting which reservations will be cancelled).
• Unsupervised learning : when the dataset assigned to the application only includes inputs , with the creation of outputs being the task to be performed in itself. In this case, the discovery of patterns in the dataset is what constitutes the outputs (e.g. from the reservations dataset construct a fixed number of customer groups – clusters – based on existing inputs , without defining any predefined -requirements).
• Reinforced learning : when the application interacts, or receives feedback for having achieved a certain objective from which it is supposed to learn (e.g. if daily the application predicts and defines that certain reservations will be canceled and these reservations end up not being cancelled, this experience should be used to improve the model).

So, in a simplistic way, the main problems to which Machine Learning is applied are:

• Classification : when the output is a discrete value, a class (e.g. the reservation will be “cancelled” or “not canceled”).
• Regression : when the output is a continuous value (e.g. what price should the accommodation be sold at).
• Grouping : when the outputs are not known and the inputs must be divided into groups (e.g. customer grouping).

There are several articles developed about how Machine Learning has already impacted or will impact the Hospitality Industry. Here are some examples (in English):

• How Machine Learning will transform the Hospitality Industry
• How Machine Learning is making hoteliers smarter
• Artificial Intelligence and Machine Learning: what will be the impact on the Tourism industry?
• Location-based machine learning gives hotels the power to match guest demands

A common aspect in most of these articles is the allusion that it is in Revenue Management that, in the short term, Machine Learning will have the most impact. Demand forecasting, revenue forecasting, customer segmentation, rate setting and booking cancellation forecasting are just some of the tasks where this impact will be noticed. However, other tasks such as predicting check-ins and check-outs per hour, employee turnover or the need for hourly employees, among many others, are also starting to be helped by Machine Learning .

At the moment, it is mostly international chains and recognized hotel brands that are taking advantage of the advantages of Machine Learning . Although many are still at an early stage of adoption, the question arises whether, like other industries, Machine Learning will become ubiquitous in all hotels. As always, the first hotels and brands to adopt it have a competitive advantage... That's why, if you're not already using it, you should at least start thinking about it.

‍

‍