Machine Learning Crash Course: The Bias-Variance Dilemma

Re:Machine Learning Crash Course: The Bias-Variance Dilemma by sgt101 on Monday July 17, 2017 @05:35PM (#14792458)

I simply don't recognise that characterisation of ML. I think that "data driven AI" fits far better. ML emerged in a number of ways over the years, but a strong driver of the last iteration was the knowledge engineering bottleneck encountered in fifth generation computing and surrounding the demise of the last turing center.

I invite you to read Chris Bishop's or Stephen Muggleton's books.

Anyone who works with data will find it hard to imagine data that "fell into your lap", all the data I've ever used successfully required slogging and grinding.

[ Parent | on HN ]

Re:Machine Learning Crash Course: The Bias-Variance Dilemma by platz on Monday July 17, 2017 @05:49PM (#14792558)

data that "fell into your lap" is a euphemism for a process in which the data was not generated by a controlled experiment in which a hypothesis is created first, and then the data is gathered according to a procedure using the proper controls and testing.

"fell into your lap" has nothing to do with how hard the work is, it's about the difference between a controlled experiment and an observational study. the bulk of ML is observational in nature (focusing on prediction) and therefore has nothing to say about causation or understanding the causal variables of the underlying reality.

[ Parent | on HN ]

Re:Machine Learning Crash Course: The Bias-Variance Dilemma by sgt101 on Tuesday July 18, 2017 @04:00AM (#14794980)

Yes, it's observational data. But this is quite common in many sciences. Observations of stella events, measurements of ecosystems and weather events for example. These have lead to theories with explanatory power and machine learning tools can and do as well.

One big deal is applications to dynamic domains.

[ Parent | on HN ]