AWS analytics tools help make sense of big data
A comprehensive collection of articles, videos and more, hand-picked by our editors
For many companies, collecting data is the easy part of the big data equation. Understanding how to process large...
By submitting your personal information, you agree that TechTarget and its partners may contact you regarding relevant content, products and special offers.
amounts of data into meaningful business decisions is where problems can arise.
In traditional big data environments, developers and data scientists create algorithms needed to make decisions. For big data on a smaller scale, a machine can effectively learn behaviors from collected historical data and compare that to current activities to make informed preventative or proactive decisions. In general, a machine learning service works through a series of processes involving modeling and comparison.
First, a mathematical algorithm is created to model the behavior of a known-good (or known-bad) condition. Then new data is collected over time and compared against that model, allowing informed decisions to be made about the new data.
A machine learning service can address three different types of tasks:
1. A binary classification model can predict one of two possible outcomes such as a yes or no response.
2. A multi-class classification model can predict multiple conditions. Multi-class classification, for example, could detect a customer's Web shopping behaviors.
3. A regression model yields an actual value or number. Regression models can predict the best selling price for a product or the number of units that will sell.
Amazon Web Services (AWS) Machine Learning service provides powerful predictive capabilities based on collected data. It can spot fraudulent transactions, predict customer behaviors or preferences, analyze unstructured documents for context, or spot impending data center equipment failures.
AWS Machine Learning offers tools, APIs and software development kits (SDKs) designed to simplify development of predictive applications. It also allows customers to create mathematical models based on historical data that can spot patterns or deviations in complex data sets. Data is stored in Amazon Simple Storage Service (S3), Amazon Redshift or MySQL databases in Amazon Relational Database Service (RDS).
How to leverage the cloud when building deep learning apps
Related Q&A from Stephen J. Bigelow
RAID 5 and RAID 6 are two types of erasure coding. The former protects data with basic parity, while the latter builds in a second layer of parity ...continue reading
Cleanly divided and straightforward applications are good candidates for a container-based deployment, whereas complex applications pose more ...continue reading
Assessing the impact of containers on application workloads can be extremely challenging, partially because of how quickly containers are spun up and...continue reading
Have a question for an expert?
Please add a title for your question
Get answers from a TechTarget expert on whatever's puzzling you.