May 11, 2021
Why Explainability Matters
When it comes to decision-making, three aspects matter most: speed, accuracy, and transparency.
Speed is a simple expression for how long it takes to make a decision. Accuracy is a measure of “making good decisions.” Transparency, on the other hand, is a far more complicated requirement of decision-making.
In many situations, we need to clearly provide an explanation for why a decision was made in a certain way. If a customer is declined credit, for example, an obscure explanation like “the computer said so” is clearly insufficient — from both a regulatory (Fair Lending) and an ethical point of view. Plus, in today’s market, borrowers have options and require more information from their banking partners. Details of the decision process are mandatory to compete.
Transparency and Explainability in Rules-Based Models
Due to lack of transparency, off-the-shelf machine learning methods – while often accurate – can be confusing and sometimes even misleading. This raises the question of which automated decision-making processes are actually suited to make high risk, critical business decisions.
If we look, as an example, at credit risk evaluations, transparency is one of the most important aspects of the decision process. Banks have to report to regulators and show that all customers are treated equally using an interpretable decision system. Customers might ask the bank for their reasoning, particularly if they are denied a line of credit. In addition, they might also wonder what to change in order to become creditworthy.
As a direct consequence from these requirements, many institutions are still using deterministic rules in order to decide whether to grant credit to a particular customer. Commonly, these rules are formulated in terms of a scorecard, or simply a set of rigid knock-out rules. The rules in the scorecard are human-readable and transparent, making it easier to explain decisions to both regulators and customers. Clearly, in terms of transparency, a decision-making algorithm comprised of deterministic rules is a great decision engine.
The Limits of Traditional Rule-Based Models
The benefits of a rule-based model that is accurate and transparent are tremendous. Rule-based models can assist in making long term business decisions as they help to assess the needs of the client base. Moreover, they provide full control over the model. Rules can be easily edited, removed, or added to a model and they make it possible to combine models of various origins.
However, rules have their disadvantages. As we know in the credit industry, rule-based systems are not very accurate. To limit the financial risks, banks put the threshold bar very high. As a result, many people are denied credit who are, in fact, good, low-risk candidates. In financial terms, this constitutes an enormous lost opportunity which is estimated to be of the order of billions of dollars in the United States alone.
A second drawback of rule-based classifiers is more subtle, but equally painful, namely scaling. When using a system of deterministic logical rules, if you need to add another rule, you must first check if the rule that is being added does not contradict any of the existing rules. For a large number of rules, this procedure becomes complicated, tedious, and costly.
Stratyfy’s Approach to Rule-Based Models: Transparency and Accuracy
Everyday, transparency in automated decision processes becomes more important, but we shouldn’t have to sacrifice accuracy. While LIME and similar technologies help to shed some light on the decisions made by black box algorithms, they yield only local explanations, and this type of explanation may be unsatisfactory or insufficient.
Stratyfy’s rule-based models, by contrast, are entirely transparent from the start to finish and provide global explanations. This makes it easy to build machine learning models based on existing systems (e.g. score cards), explain decisions to clients and regulators, and to develop human-interpretable models for marketing and research.
How Stratyfy’s Solution Works
Stratyfy mimics the high-level human decision process, weighing factors with various degrees of importance. By then rule mining to supplement the engine with additional rules, we can create models that are even superior in terms of accuracy and can even achieve levels of accuracy on par with black box approaches. Stratyfy even allows for rules specified directly by humans – experts representing their knowledge of the field.
The result?
- Each individual factor that contributed to the decision is human-interpretable.
- You can assess in which way each factor contributed to the decision. In particular in the case of a yes/no decision, you can understand whether a factor was in favor or against, and formulate a measure to determine how important the factor was in the decision process.
- The model of decision is simpler than a very complex, black box model – which means you don’t need a data science team to leverage advanced analytics.
The typical workflow to work with Stratyfy’s Rule Engine is:
- Specify a set of rules (“expert knowledge”) — optional
- Provide reference data as training set — optional
- Calibrate the Rule Engine and evaluate performance
- Use Rule Mining to find additional rules
- Calibrate Rule Engine on the new set of rules and evaluate performance
- Use calibrated model for default prediction on new data
As a result, we obtain a decision model consisting of rules and their weights. This resulting model is entirely transparent and interpretable – a first of its kind in our industry. For more, read the full white papers: part 1 and part 2.
If you’d like to learn more about how Stratyfy can help you develop transparent rule-based models, contact us.