Problem 24: Can Data Predict the Future?

Machine Learning

Concepts appear as they come up…

So far, we've been asking questions

00:00/01:32

1 / 1

Machine Learning

Certisured is an Edtech based out of Bangalore and is Co-founding team of Wizori - built to provide quick fundamental clarity in form of short vibes.

www.certisured.com

Machine Learning

Loading comments…

Continue your learning — your way

Sign in to unlock the full playlist, save your progress, and get a personalized feed.

Up next

Problem 6: Do We Really Need a Server?
1:29
Problem 22: How Do We Analyze Massive Data?
1:39
Problem 4: Who Should Handle the Next User?
1:32
Problem 5: what if demand for machines increases?
1:26
Problem 7: Are Files Enough for Business?
1:45
Problem 2 : Computer is temporary but, Data is permanent
1:20
Problem 21: Where Should We Keep Our Secrets?
1:44
Every Cloud Service Begins With a Problem
1:16
Problem 20: How Do Services Complete One Workflow?
1:38
Problem 3: Shared Object Storage
1:32
Engineering Problem 1: Renting a Computer
1:23
Problem 8: Must We Ask the Database Every Time?
1:38

Problem 24: Can Data Predict the Future? — full transcript

Machine Learning

Machine Learning

So far, we've been asking questions about our data. How many products were sold? Which city generated the highest revenue? How many users logged in yesterday? These questions explain the past. But businesses wanted something more. Can we predict which customer is likely to leave? Can we detect fraud before it happens? Can we recommend the next product a customer might buy? The engineering problem became clear. How do we build systems that learn patterns from data instead of following fixed rules? The engineering concept that solved this problem is Machine Learning. Instead of writing rules for every situation, engineers train a model using historical data. The model discovers patterns, learns relationships and uses that knowledge to make predictions on new data. As more data becomes available, the model can be retrained and improved. Amazon Web Services provides this through Amazon SageMaker. Microsoft Azure provides Azure Machine Learning. Google Cloud provides Vertex AI. Different names. One engineering concept. Applications no longer just stored and analyzed data. They began making intelligent predictions from it. But another challenge soon appeared. Building a machine learning model is one thing. How do we train models that require enormous computing power without managing the underlying infrastructure ourselves?