{"id":943,"date":"2018-10-25T03:25:09","date_gmt":"2018-10-25T00:25:09","guid":{"rendered":"http:\/\/kusuaks7\/?p=548"},"modified":"2021-05-11T14:00:41","modified_gmt":"2021-05-11T14:00:41","slug":"the-machine-learning-workflow","status":"publish","type":"post","link":"https:\/\/www.experfy.com\/blog\/ai-ml\/the-machine-learning-workflow\/","title":{"rendered":"The Machine Learning Workflow"},"content":{"rendered":"

Ready to learn Machine Learning? Browse<\/em><\/strong> Machine Learning Training and Certification courses<\/a> developed by industry thought leaders and Experfy in Harvard Innovation Lab.<\/em><\/strong><\/p>\n

What's different about machine learning projects? How do you reduce risks and build a good solution quickly?<\/h4>\n

 <\/h4>\n

\"image\"<\/p>\n

Machine Learning Project Workflow<\/h4>\n

In standard software development, you simply answer the question:<\/p>\n

\n

What do you want to implement?<\/em><\/p>\n<\/blockquote>\n

And then you, well, implement.<\/p>\n

But in machine learning projects, you first need to explore<\/strong> what’s possible – with the data you have. So the first question is:<\/p>\n

\n

What can you<\/strong> implement?<\/em><\/p>\n<\/blockquote>\n

\"Machine<\/p>\n

Here’s what we learned works to keep a machine learning project on track from start to finish:<\/p>\n

1. Define the task<\/h2>\n

\"Step<\/p>\n

It’s easy to get drawn into AI projects that don’t go anywhere. A proper machine learning project definition drastically reduces this risk.<\/p>\n

These are the questions you need to answer to define a project:<\/p>\n

Understand the current process<\/h3>\n

What is your current process? Your machine learning solution will replace a process that already exists. How are decisions currently made in this process? Considering the current process will give you a lot of domain knowledge<\/em> and help you define how your machine learning system has to look.<\/p>\n

Define what you want to predict<\/h3>\n

What exact variable do you want to predict? Define the output of your machine learning system — in as much detail as possible.<\/p>\n

List the useful data sources<\/h3>\n

What data do you have that’s useful<\/strong> to predict this output correctly? Start by listing the data sources the current process relies on. One way to list useful data sources is by asking yourself: “If I — as a human — needed to make this prediction, what data points would I want to know about?<\/em>”<\/p>\n

If you understand the current process, know what you want to predict, and have identified all the useful data sources, then you’re in a good position to decide whether it makes sense to proceed to the next stage.<\/p>\n

2. Find an approach that works<\/h2>\n

\"Step<\/p>\n

Even if you have a good problem definition, you can’t know yet how accurate your machine learning model will be in the end – or whether it will be worth replacing the current process.<\/p>\n

A proof of concept is the cheapest way to find out what ROI you can expect from your final solution. These are the steps:<\/p>\n

Research<\/h3>\n

Research all the ways other teams have resolved similar tasks — whether they used machine learning or not. Then make a plan<\/strong>, using what you’ve learned from both your research and the existing process you want to replace.<\/p>\n

Build a dataset<\/h3>\n

The central part of any machine learning project is the sample dataset! This includes realistic <\/em>examples of exactly those cases for which you want your machine learning system to make correct predictions. Think of it as an Excel table, with:<\/p>\n