inputcoffee

9 ms·

91.
by inputcoffee 9y ago
I suggest that the undergrads be encouraged to go to Arxiv and browse papers, try to work through them and see the range of papers. The reason for this is that many papers aren't all that well written and well argued. Many are, to be s
92.
by inputcoffee 9y ago
https://amp.azcentral.com/amp/74412576
93.
by inputcoffee 9y ago
You pump the water from the sea. With free energy. I’m not the first to point this out. There are pages on it. Please google it.
94.
by inputcoffee 9y ago
Desalination
95.
by inputcoffee 9y ago
They have some of the best solar irradiance. Free energy would solve any water problems.
96.
by inputcoffee 9y ago
Sidewalk Labs is making Toronto a smart city, and Amazon is presumably going to develop a second HQ. three investment opportunities, folks
97.
by inputcoffee 9y ago
Yes, the latter.
98.
by inputcoffee 9y ago
One of the cognitive science labs was hiring experts in graph theory, and it turns out they wanted to express Arrow's Theorem -- and the related theories -- in the form of a graph so they could reason about decision making in the brain
99.
by inputcoffee 9y ago
I misread it as "to study that at college" instead of "study at that college"! I now have no idea why it was down-voted.
100.
by inputcoffee 9y ago
So true. I have had to do this. I had a pretty good regression model but it was not taken seriously. So I wrote it using "a neural network in TensorFlow" and the next thing you know the whole company is asking me how it works and
101.
by inputcoffee 9y ago
The XOR problem is famously solvable by adding a layer to a single layer perceptron assuming a unit step function. This is a very basic exercise taught in many intro courses. I agree with every second sentence: Given that every perceptron l
102.
by inputcoffee 9y ago
Wealth could be defined as consumption. (Not necessarily in the sense of using up resources. You could include externalities). Winning is defined as increasing consumption. Everyone can have access to more of something. For instance, there
103.
by inputcoffee 9y ago
Wait a second, a 2-layer perceptron can solve the XOR problem. The first layer maps it to a linearly separable problem which the second layer solves. Furthermore, a non-linear 2-layer node can map any arbitrary function so the argument fo
104.
by inputcoffee 9y ago
I didn't down-vote you but I suspect people reacted negatively because the implication is you have to stop learning after college. They may feel you can always learn it at any time. You can take deep learning and math courses online et
105.
by inputcoffee 9y ago
I wouldn't consider it "a problem," but a theory is surely better than the lack of one. It may be out of our league, or we may just need a genius to help figure it out. Calculus and Newtonian Mechanics and Relativity were pro
106.
by inputcoffee 9y ago
When I wrote that, I was thinking of the unit step function, for which the derivative is not defined.
107.
by inputcoffee 9y ago
I didn't mean to say that it is the only problem. It is more like the "meta problem" as far as theory is concerned. There are many parameters (layers, nodes, connectivity, drop off, number of trials etc). Right now they are g
108.
by inputcoffee 9y ago
I feel like asking: did they solve the problem? Let me see if I can state the problem: Neural Networks are non-linear because of their activation functions. You need a differentiable function in order to take the derivative so you can back-
109.
by inputcoffee 9y ago
You state it much more clearly than the author does. While I would not go so far as to call it "silly". I think an interesting point is obscured for the purpose of a dramatic headline.
110.
by inputcoffee 9y ago
A date is crucial, even for a programming related task. There are probably people who, at this very moment, are looking up Django 1.0 documentation right now and can't get the example to work because their system installs the second to
111.
by inputcoffee 9y ago
This article had two components that I appreciate: 1. It gave us a broad overview of the field with data (example, the relative performance on various algorithms on certain problems) 2. It gave us another way of looking at some math we may
112.
by inputcoffee 9y ago
Two important points: 1. What about wind and solar? 2. The death/unit energy misses out the fact that we spend a lot more to keep nuclear safe because we are worried about it. If we spent a fraction of the same amount on other energy,
113.
by inputcoffee 9y ago
I am very curious how you teach coding on a mobile platform. I mean there is so much typing in coding, and then you need to track a bunch of variables, see the output, and see the code, so that is a lot of screen real estate you need.
114.
by inputcoffee 9y ago
Agreed. I would just like the OP but since you can't see the like count, it is valuable feedback to see how many people find the name confusing. I also assumed it was python focused.
115.
by inputcoffee 9y ago
Can only one person "apply" to solve it? I'm thinking of the kickstarter problem: vaporware. If, over 6 months, 5 people apply to solve a problem, who gets the subscribers? The first one, I hope not. Each subscriber chooses?
116.
by inputcoffee 9y ago
Not sure why you are getting down voted but I was thinking along the lines of the healthcare point. It is a Nassim Taleb type of point: you're basically making extra money writing healthcare puts, but when there is a crash, you get wip
117.
by inputcoffee 9y ago
I always use the example of Japan. $1 = 109 Yen, so do people think Japan costs 1/100th of what the US does? Japan is a good example because people have an idea that it is a rich, expensive country.
118.
by inputcoffee 9y ago
Just say "you look at the manhole covers and ask 'why'; I look at the same covers and ask 'why not'" That's the secret answer at McKinsey.
119.
by inputcoffee 9y ago
Perhaps you are not acquainted with the severe name shortage that currently threatens the Javascript community? https://github.com/orbitaljs https://www.npmjs.com/package/conker https://www.
120.
by inputcoffee 9y ago
The big question in data science is: should I spend more time learning Python or R? The answer is always: math

More ›