Machine Learning Is Coming to the Mainframe

Don’t blink! Another major technology revolution is upon you, and it’s getting everyone’s attention: Machine Learning and advanced analytics come to the mainframe! CA Technologies recently unveiled new machine learning capabilities that are a comprehensive strategy for the efficient management of the mainframe production environment.

“So what?” you might ask. Machine Learning and intelligent mainframe operations are your keys to driving MTTR toward zero. With it, you can empower your IT team to leverage predictive analytics and automated remediation to reduce MTTR by fixing problems before they impact the business.

What Exactly Is Machine Learning?

Machine Learning (ML) has been here for decades, and many of the concepts and algorithms have been around for many years. It is exciting technology that allows processing enormous amounts of data. Thanks to today’s performance optimized hardware, this data can be analyzed within an unbelievably short amount of time.

ML itself is an interdisciplinary field that shares many paradigms and concepts with other fields of mathematics and statistics and can be also viewed as a part of Artificial Intelligence (AI). However, in contrast to AI, ML does not try to imitate an intelligent behavior, but rather focuses on algorithms that can process huge volumes of data and detect patterns that are not obvious or easily deduced by humans. For example:

• ML algorithms create “models” based on some known data. Models then make data-driven predictions (decisions) on new, unseen data. It means that the model is not a program you would code but instead is generated logic that can interpret the data and provide some output.
• The algorithms learn from and make predictions on data. So, once again, “it’s all about the data” – data are crucial for starting and continuing ML. 

Machine Learning Applicability

ML can be helpful in the following areas:

1. Classification. Example: A spam filter could be a good candidate, or any other category or class assignments.
2. Regression. Like classification, but the output is not a category or class. Example: Temperature forecasts and stock price changes.
3. Clustering. Often complements data mining, is helpful when you want to find some data patterns. Example: Anomaly detection.

It’s All About the Data

You can’t just pick up a Machine Learning tool, feed it data, and magically receive perfect output. There is no “silver bullet” for today’s data scientists (I believe that this will be one of the key roles in the coming millennium). Some basic “block and tackle” is needed that should include:

• Review the data.
• Extract the valid properties.
• Convert and transform the data (if needed).
• Select a relevant model.
• Train, validate, and deploy.

So, Where to Start?

By now, you probably can see I am a Machine Learning enthusiast who loves to share my knowledge. I’d like to give you a few points of reference where you can begin your own research and hopefully develop a love for ML as I have. Please consider these:

• Books
  • Elements of Statistical Learning by Hastie, Tibshirani, and Friedman.
  • Pattern Recognition and Machine Learning by Christopher Bishop.
  • Understanding Machine Learning: From Theory to Algorithms by Shai Shalev-Shwartz and Shai Ben-David.
• Online courses and YouTube
  • Coursera.
  • IBM Big Data University.
  • Khan Academy.
• Tools
  • R: The “gold standard” in statistics, ML, and other fields including data visualization.
  • Python. Another very popular language. Libraries like pandas or sckit-learn provide Machine Learning algorithms for Python.
  • Apache Spark is a relatively new kid on the block. Spark is a very efficient and highly parallelizable framework and a Machine Learning library is one of its core components that is included in the base install.
  • Other tools are available at no cost (yes, free!). Just to name a few: Weka, Julia language and Machine Learning packages, and Google’s TensorFlow. 

Machine Learning Keep on Learning

What I’ve shared with you in this blog is just the “tip of the iceberg” and I hope it gave you an idea of what ML can provide to you. For now, if you are excited to get started, think about your data and possible use cases. Then, learn more about ML from the sources mentioned earlier.

I’d like to keep the conversation going, so if you have any ML tips, tricks or best practices, please share. I’d also like to hear from you if you’re interested in more, deeper, ML discussions, and applicability.