Estimated reading time: 6 minutes
Companies are generating ever-increasing amounts of data associated with business operations, leading to renewed interest in predictive analytics, a field that analyzes large data sets to identify patterns, predict outcomes and guide decision making. Companies also face a complex and ever-expanding range of operational risks that need to be proactively identified and mitigated. While many companies have begun using predictive analytics to identify marketing/sales opportunities, similar strategies are less common in risk management, including security.
Classification algorithms, a general class of predictive analytics, could be particularly useful for the refining and petrochemical industries by predicting the timing and location of safety incidents based on safety-related inspection and maintenance data, essentially leading indicators. There are two main challenges associated with this method: (1) ensuring that the measured leading indicators are actually predictive of crashes and (2) measuring the leading indicators frequently enough to have predictive value.
Using regularly updated inspection data, a model can be created using a logistic regression. This way you could create a model, for example, to predict the probability of rail failure for each mile of track. Probabilities may be updated as additional data is collected.
In addition to the predicted probabilities of rail failure, with the same model we can identify the variables with greater predictive validity (those that significantly contribute to rail failure). Using the model results, you will be able to identify exactly where to focus maintenance, inspection and capital improvement resources and what factors to address during these activities.
The same methodology could be used in the refining and petrochemical industries to manage risks by predicting and preventing accidents, provided that organizations:
Predictive analytics is a broad field that encompasses aspects of various disciplines, including machine learning,artificial intelligence, statistics and data mining. Predictive analytics uncovers patterns and trends in large data sets. One type of predictive analytics, classification algorithms, could be particularly beneficial to the refining and petrochemical industries.
Classification algorithms can be classified as supervised machine learning. With supervised learning, the user has a dataset that includes measurements of predictive variables that can be linked to known outcomes. In the model discussed in the case study section of this article, various track measurements (e.g. curvature, crossings) were taken during a period for each mile of track. The known outcome, in this case, is whether a track failure occurred on each rail mile during that two-year period.
An appropriate modeling algorithm is then selected and used to analyze the data and identify relationships between variable measurements and outcomes to create predictive rules (a model). Once created, the model is given a new dataset containing measurements of unknown predictor variables and outcomes and will then calculate the probability of the outcome based on the model's rules. This is compared to types of unsupervised learning, where algorithms detect patterns and trends in a dataset without any specific direction from the user, other than the algorithm used.
Common classification algorithms include linear regression, logistic regression, decision tree, neural network, support vector/flexible discriminant machine, naive Bayes classifier, and many others. Linear regressions provide a simple example of how a classification algorithm works. In a linear regression, a best fit line is calculated based on the existing data points, giving the line equation ay = mx + b. Entering the known variable (x) provides a prediction for the unknown variable (y).
Most relationships between variables in the real world are not linear, but complex and irregularly shaped. Therefore, linear regression is often not useful. Other classification algorithms are capable of modeling more complex relationships, such as curvilinear or logarithmic relationships. For example, a logistic regression algorithm can model complex relationships, can incorporate non-numeric variables (e.g., categories), and can often create realistic and statistically valid models. The typical output of a logistic regression model is the predicted probability of the outcome/event occurring. Other classification algorithms provide similar output to logistic regression, but the required inputs are different between algorithms.
Modeling complex relationships is particularly useful in risk management, where risk is typically prioritized based on the likelihood and potential severity of a particular outcome. Modeling the risk factors that contribute to that outcome results in a precise and statistically valid estimate of the probability of the outcome. In contrast, many risk assessments measure “probability” on a categorical scale (once a decade, once a year, several times a year), which is less precise, more subjective, and makes it impossible to distinguish between risks present in the risk. same broad category. There are other techniques for quantifiably assessing potential severity in a risk assessment, but this is beyond the scope of this article.
BlogInnovazione.it
Developing fine motor skills through coloring prepares children for more complex skills like writing. To color…
The naval sector is a true global economic power, which has navigated towards a 150 billion market...
Last Monday, the Financial Times announced a deal with OpenAI. FT licenses its world-class journalism…
Millions of people pay for streaming services, paying monthly subscription fees. It is common opinion that you…
