Machine Learning within the Travel Money Industry

In recent years, machine learning’s irrefutable benefits have attracted the attention of some of the biggest organizations.

The objective of machine learning is to remove the need for human capital within data-driven processes. A machine will no longer require human capital to input learned experiences and make a model more accurate. Instead, a machine will learn from itself.

There are four benefits which can significantly transform your business’ profitability by reallocating or removing the need for human intervention in traditional processes:

Better Value

Currently the travel money industry is very labor intensive. With the minimum wage destined to increase, the costs savings from reducing or reallocating human resource is growing.

More Accurate

Inherently, individuals are subject to their own biases and operate with errors. Machines that do not require human intervention do not suffer from either of these two disadvantages, thereby increasing the model’s accuracy.


Any process loop can be made shorter by reducing the number of steps. Subsequently, by making the loop shorter we can increase the speed of the process. This allows the machine to react faster to market changes; something which is seen by the travel money industry as vital to protecting profitability.

Greater Capacity

Machines have far greater capacity than individuals, this enables a more holistic view determined from a greater array of inputs.


Within the field of machine learning, there are three core types: supervised, unsupervised and reinforcement learning. Each of the three types has a specific area or operational task they are more equipped to deal with.

Supervised learning

Supervised learning maps inputs to outputs creating an inferred function ideal for making predictions and drawing together classifications.

Unsupervised learning

Unsupervised learning also maps inputs to outputs however the data it processes is unlabelled. By utilising unlabelled data, unsupervised learning can isolate previously unseen patterns and clusters within the data.

Reinforcement learning

Reinforcement learning optimises actions to maximise a cumulative reward. The model interacts with an environment and from that interaction receives two key metrics as an output, ‘the state’ and ‘the reward.’

All three types of machine learning have a place within the travel money industry. However, reinforcement learning is particularly well suited to developing a dynamic pricing model. Its ability to react to market changes and remove the cost of human capital makes it the most optimal strategy to implement.

CMS Analytics is working with travel money operators to help implement machine learning – particularly with pricing. The white paper, ‘Machine Learning: Leveraging Dynamic Pricing to Achieve Strategic Goals’, provides greater insight and worked examples of this.

Get In Touch

What do you want to speak with us about?

Tel: 312 288-8413   |

CMS Analytics, 55 Allen Plaza, Suite 320, 55 Ivan Allen Jr. Boulevard, Atlanta, GA 30308

Cash Management Systems Limited (3128552 England). Copyright © 2019 Cash Management Systems Limited.