Weka

See the following -

Artificial Intelligence and Machine Learning Bias has Dangerous Implications

Algorithms are everywhere in our world, and so is bias. From social media news feeds to streaming service recommendations to online shopping, computer algorithms—specifically, machine learning algorithms—have permeated our day-to-day world. As for bias, we need only examine the 2016 American election to understand how deeply—both implicitly and explicitly—it permeates our society as well. What’s often overlooked, however, is the intersection between these two: bias in computer algorithms themselves. Contrary to what many of us might think, technology is not objective...

Machine Learning in Healthcare: Part 2 - Tools Available to the Average Healthcare Worker

A variety of machine learning tools are now available that can be part of the armamentarium of many industries, to include healthcare. Users can choose from commercial expensive applications such as Microsoft Azure Machine Learning Studio, SAS Artificial Intelligence Solutions or IBM SPSS Modeler. Academic medical centers and universities commonly have licenses for commercial statistical/machine learning packages so this may be their best choice. The purpose of this article is to discuss several free open source programs that should be of interest to anyone trying to learn more about machine learning, without the need to know a programming language or higher math.

Read More »

Machine Learning in Healthcare: Part 3 - Time for a Hands-On Test

Every inpatient and outpatient EHR could theoretically be integrated with a machine learning platform to generate predictions, in order to alert clinicians about important events such as sepsis, pulmonary emboli, etc. This approach may become essential when genetic information is also included in the EHR which would mandate more advanced computation. However, using machine learning and artificial intelligence (AI) in every EHR will be a significant undertaking because not only do subject matter experts and data scientists need to create and validate the models, they must be re-tested over time and tested in a variety of patient populations. Models could change over time and might not work well in every healthcare system. Moreover, the predictive performance must be clinically, and not just statistically significant, otherwise, they will be another source of “alert fatigue.”

Read More »