{"id":24887,"date":"2019-12-02T09:03:03","date_gmt":"2019-12-02T14:03:03","guid":{"rendered":"https:\/\/intelligencecommunitynews.com\/?p=24887"},"modified":"2019-12-02T09:03:09","modified_gmt":"2019-12-02T14:03:09","slug":"your-piecemeal-approach-to-machine-learning-is-all-wrong","status":"publish","type":"post","link":"https:\/\/intelligencecommunitynews.com\/your-piecemeal-approach-to-machine-learning-is-all-wrong\/","title":{"rendered":"Your piecemeal approach to machine learning is all wrong"},"content":{"rendered":"\n

Deploy machine learning at the enterprise level in your agency by combining it with DevOps<\/em><\/p>\n\n\n\n

By Chris Powell, CTO of NT Concepts<\/p>\n\n\n\n

Bottom line upfront<\/h1>\n\n\n\n

Machine learning is a game-changer for the\nIntelligence Community. Unfortunately, many agencies only apply machine\nlearning solutions to narrow challenges. The result? They see one-off successes\nin niche categories, but these wins are not tied to any broader goals or even\nto each other. Agencies fail to reap enterprise-wide benefits.<\/p>\n\n\n\n

Specifically, machine learning models can\n\u201clearn\u201d on their own based on the continuous evaluation of data. This automated\napproach to analyzing data frees up your agency employees to perform more high-value\ncreative analysis.<\/p>\n\n\n\n

But to get to this point\u2013\u2013leaving the\nmachines to do the grunt work and your people to do the visionary work\u2013\u2013your\nagency needs to approach machine learning from the enterprise standpoint.\nHere\u2019s how.<\/p>\n\n\n\n

Combine Machine Learning with DevOps<\/h1>\n\n\n\n

When you look across agencies, there are\ndozens of different tools and applications to complete missions. The abundance\nof data is voluminous. In fact, it staggers many agencies; they simply don\u2019t\nknow how to put it to practical use.<\/p>\n\n\n\n

It\u2019s our strong assertion that machine\nlearning can\u2019t be a standalone data science exercise. It needs to be baked into\nyour DevOps (i.e., your software development and IT operations). This combined\napproach materially moves the ball forward for your agency.<\/p>\n\n\n\n

How? Your systems and operations are linked\nto machine learning models\u2013\u2013continually feeding the models new data that they\ncontinuously evaluate and re-evaluate to learn more and to learn better,\nproviding your analysts and data scientists with the best information. In\nshort, it\u2019s a never-ending successful feedback loop.<\/p>\n\n\n\n

This is called Machine Learning Operations\n(MLOps). Let\u2019s take a closer look.<\/p>\n\n\n\n

Set up an\norchestrated pipeline for data<\/h1>\n\n\n\n

Right now, your\nagency is likely set up with a machine learning pipeline. A narrow dataset\nflows into a machine learning model designed to solve a narrow challenge. Yet\nyour software deployment has a pipeline of data, too, as well as your software\nintegration. As a result, you have multiple pipelines of data flowing through\nyour agency.<\/p>\n\n\n\n

To achieve maximum\nvalue of machine learning at an enterprise level, all your machine learning and\nDevOps pipelines of data need to converge into one MLOps pipeline.<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

Machine Learning Operations (MLOps) Pipeline<\/em><\/strong><\/p>\n\n\n\n

The MLOps pipeline\nprovides your team with:<\/p>\n\n\n\n