In response to an upcoming UCSD seminar titled Knowledge Graph Construction using Machine Reading Methods I summarized an idea on the nature of technical knowledge.
My analysis of technical knowledge in manufacturing, aviation, and elsewhere suggests that it is fractal, i.e. that any portion of a knowledge graph can be further decomposed into a detailed knowledge graph in its own right. Limits on human knowledge mean that the frontiers of current graphs are always “fuzzy,” i.e. at low stages of knowledge. Further technology development will clarify clarify the current periphery of a graph, but reveals new fuzzy portions.
The seminar is by Ndapa Nakashole (CMU), Feb. 8. I’m hoping to discuss whether she has run into anything similar in natural language processing.
To the extent this hypothesis is true, i.e. that knowledge is fractal, it has a lot of implications. For example, high-tech industries must operate in frontier regions where much is known, but some important issues are not well understood. People are better than machines at dealing with ambiguity, so the faster the rate of technological progress, the more an industry needs people and cannot automate its activities.
Here’s a column by a Forbes blogger about Zika saying that “we should not wait so long to develop vaccines against tropical diseases.” He concludes:
Many pharmaceutical companies don’t focus on a disease until it becomes common enough to be highly profitable. The trouble is the vaccine world has become a bit like the plot line for “She’s All That” or “Cinderella.” Attention towards a person or thing does not occur until a cool person notices he or she or it. But when it comes to disease and stock market opportunities, as the saying goes, once your grandmother knows about it, it is usually too late.
Source: Zika Vaccine: Another Example Of Waiting Until It’s Too Late? – Forbes
This is not news. And it’s a classic situation where market forces are not enough to give socially desirable behavior. Developing a vaccine for a disease that is not in rich countries has low expected profitability. Even if the disease goes epidemic, pharma company will have to sell at a price near marginal cost.
The only solution is to use a different way to fund development. Contests, grants (Gates foundation), purchase guarantees (used by US DoD) all work. But waiting for the traditional patent system + pharma profit motive won’t lead to timely development of medication for poor-country diseases.
I guess a Forbes columnist is not allowed to point this out.
What’s a good place to put supplemental information, especially photos and tables, for my book? I have a lot of old photographs, and putting them into the book itself gets expensive. Some are in color and some are very large. Here are a few examples.
I could set up my own site, or use my publisher’s, but places like Tumblr know how to run photo sites. The ideal features I want include being able to link to pictures on other sites (due to copyright restrictions), able to create tables of contents, etc. Straight chronology won’t suffice.
include Tumblr, Pinterest, Instagram. I don’t use any of them except to dabble, so I don’t know their strengths. Possibly Twitter or Facebook?
All advice welcome. Email me, or post comments here.
Wired has a good article on fires from “hoverboards,” which are essentially very small hands-free electric scooters. Here is an example. Start at about the :30 mark to see how these devices explode, and can easily set a house on fire.
nor other news stories have much constructive advice about how to reduce the chance of fires. As they point out, buying a well-made model is important, but at present there is no way to distinguish the well-made ones from the knock-offs. And most are
cheap knock-offs by companies that will be gone in a year.
I have some experience with the underlying cause of these fires, their Lithium-based batteries, because I use them in radio controlled aircraft. Fires of these batteries are not common, but they happen. Two people in San Diego who I know directly have had major fires. One lost a 2-unit condo, the other a detached workshop. The second one happened to an expert in RC flying!
One contributor to the A320 crash off Brazil in 2009 (Air France 447) was that the two pilots were making opposite inputs on their control sticks. The aircraft was in a stall, and therefore it was crucial to push the nose down, to regain airspeed. The instinctive human reaction (of untrained people) is to pull the nose up, since the airplane is falling. To oversimplify a long sequence of events drastically, pilot made the correct move, but the other pilot apparently panicked, and pulled back on his control stick. He continued to do this as they fell from 40,000 feet all the way to the Atlantic Ocean.
A new accident report says that the same thing happened in the crash of an Indonesia AirAsia Airbus A320, flight QZ8501, last year.
I was teaching the Virginia Mason VMMC case in Tech & Operations Management yesterday, and made a loose comment about busy urban hospitals being better than suburban ones. For example in the UC San Diego system, when someone is my family is really sick I try to take them to the downtown (dilapidated, overcrowded) UCSD hospital before I’d go to the one near campus (hotel-like, luxurious).
A student asked “why”, forcing me to do a little research. Here is my answer to her. Continue reading
When faced with lots of variables and likely interaction terms, is linear regression usable? In the comments, I get educated that yes, it can be done. Thanks to Nigel Goodwin. I will still tell my students to try some inherently nonlinear methods, though.
Source: Logistic Regression Vs Decision Trees Vs SVM: Part I