Just because I missed posting for the last year, doesn’t mean I have not been writing. Perhaps I have been writing more. Here is something that I just wrote for a perspective on opportunities for machine learning in population health.
Machine learning (ML) is emerging as a technology, climbing the “peak of inflated expectations” or perhaps even starting to slip into the “trough of disillusionment”, in the terms of the technology hype cycle,[ref] and offers both opportunities and threats to population health. ML is a technique for constructing computer algorithms, and what distinguishes ML methods from other computer solutions is that, while the structure of the computer program may be fixed, the details are learned from data. This data-driven approach is now dominant in Artificial Intelligence (AI), especially through deep neural networks, and stands in contrast to the old way, an expert-algorithms approach in which rules summarizing expert knowledge were painstakingly constructed by engineers and domain specialists. ML has succeeded by trading experts and programmers for data and nonparametric statistical models. However, the applications where ML has been successfully deployed remain limited. AI luminary Andrew Ng provides this concise heuristic: “[i]f a typical person can do a mental task with less than one second of thought, we can probably automate it using AI either now or in the near future.”[ref]
The editor only wants 1,000 words, so this is getting cut.
I was nearly convinced that Google’s TensorFlow would take over the world, but now I’ll need to also consider MXNet: http://mxnet.io/
A short paper from my work on predicting who will get elective surgery for diverticulitis is on arXiv: https://arxiv.org/abs/1612.00516 [tag: my-research]
From the sklearn docs: “We call our estimator instance `clf`, as it is a classifier.” http://scikit-learn.org/stable/tutorial/basic/tutorial.html#learning-and-predicting
Darcy AM, Louie AK, Roberts L. Machine Learning and the Profession of Medicine. JAMA. 2016;315(6):551-552. doi:10.1001/jama.2015.18421.
> Must a physician be human? …
I have been getting some great success from the scikits-learn CountVectorizer transformations. Here are some notes on how I like to use it:
ngram_range = (1,2)
clf = sklearn.feature_extraction.text.CountVectorizer(
min_df=10, # minimum number of docs that must contain n-gram to include as a column
#tokenizer=lambda x: [x_i.strip() for x_i in x.split()] # keep '*' characters as tokens
There is a
stop_words parameter that is also sometimes useful.
EnsembleMatrix: Interactive Visualization to Support Machine Learning with Multiple Classifiers http://research.microsoft.com/en-us/um/redmond/groups/cue/publications/CHI2009-EnsembleMatrix.pdf
I want one
It was great speaking with you. This is the paper I was talking about.
Looking forward to know more about each other’s work.