Follow Slashdot stories on Twitter


Forgot your password?

Comment The stuff is just too expensive (Score 5, Insightful) 84

I think what will kill iot is that it's just frankly too expensive. A perfect example is the Belkin WeMo line of iot enabled products.

* 150 dollars for a slow cooker

* 150 dollars for a coffee maker

* 200 dollars for a humidifier

* 40 dollars for a plugin relay switch

And the list goes on. The nest costs 5x-10x more than a low end digital thermostat. I have a sneaking suspicion as with almost all other home automation, upper class people will buy it for the novelty but the rest of the world will keep to their "dumb" devices.

Comment Re:I can understand small first batches (Score 2) 109

Unfortunately the pi draws a fair amount of current doing nothing. I've used the pi in a solar based project and the pi ended up using more energy than my solenoid. I ended up using an Arduino and a relay to turn the pi off to conserve energy. It also made the project somewhat complicated because you can't just power off the pi, you have to do a proper linux shutdown. You have to coordinate with the AVR to say "OK to kill power now." You're typically looking at about 100mA to 300mA depending the model. The AVR uses less than 1mA in sleep mode.

Comment Re:I can understand small first batches (Score 5, Interesting) 109

The ESP8266 seems to be the current competitor for "arduino with effortless network connectivity." They are about 5 dollars and actually available. Just as the pi, it has many limitations that dedicated microcontrollers solved years ago. I've been using a combination of ESP8266 and AVR lately instead of ESP8266 standalone.

Comment Re:I can understand small first batches (Score 5, Informative) 109

The zero is definitely in an awkward spot. It's so-so as a microcontroller replacement (no low power modes, limited number of analog inputs, inputs not 5v tolerant), but the price point makes it otherwise competitive. The ESP8266 is getting high level languages like Lua, micro python, and Basic and priced well. The "low power mode" sucks because it basically just resets the unit and doesn't have interrupt driven wake modes.

I bought a zero and it will probably be my only purchase. It just doesn't seem to do anything special.

Comment Re:the Schedules (Score 2) 109

I've noticed Adafruit runs out of the zero within minutes of stock, but the more expensive "kits" may last several hours. I think it really comes down to the 5 dollar price point not being profitable and thus not being a priority. I'm sure at 35 dollars the B+ has at least _some_ wiggle room.

Comment Re: Militant Slashdot (Score 3, Insightful) 293

I frankly do not understand gun control in America. Gun control seems to boil down to
1) Getting rid of "assault rifles"
2) Consistent background checks
3) Magazine sizes

That's great and all, but the vast majority of gun violence are handguns. Even more, the .22lr of all things seems to be the deadliest caliber. Whether or not you are for gun control, let's discuss the actual killer: handguns. All this other stuff is just a distraction.

Comment Saying "AI" will get you labelled a quack (Score 1) 149

I have some involvement in this field and I can't think of a single time I've seen the term AI in a book or research paper. The only time I ever see anyone use it is in the media or various futurists. Usually people just name their specific subfield or name the types or algorithms in which they specialize. "I specialize in unstable learners" or "I specialize in transfer learning" but I've never heard someone say "I specialize in AI."

I think naming neural networks as they did was probably a very bad idea. The mathematics have very little to do with actual neurons in the brain but the name leads you to believe differently.

Comment Re:Embrace (Score 2) 105

Microsoft actually has no other product like R, so I would imagine the real reason they bought R was so they could add an additional layer to their data analysis tools inside SQL server. It's actually a no-brainer. Do a SQL server query, store results in a data frame, run a clustering algorithm against the data to see purchasing habits.

Comment R is the wild wild west (Score 1) 105

The R ecosystem seems to be the opposite of Microsoft's traditional ecosystems. About the only thing R library designers can agree upon is "dataframe is good." Packages that try to put a consistent front-end on other packages (i.e. caret) definitely helps. However, even something as simple as "does this algorithm want a factor or dummy variables?" may require examining the source code. Other more subtle things like "Does it was the data to be centered and scaled?" may slip by.

I hope Microsoft addresses this. As a researcher, a common task is to compare the performance of many algorithms against a new dataset or a new algorithm that you are developing, and it can be a pain to do in R. Weka is WAY more consistent in handling these sorts of things, but Weka only handles a subset of the tasks people do in R. Something as simple as style-guide docs would go a LONG way.

Comment Re:R vs. Python vs. other (Score 5, Informative) 105

R is preferred by statisticians. Many statisticians are on the leading front of creating new traditional machine learning algorithms (not the GPU driven or map reduce stuff "hip" companies are dealing with right now). Things like supervised classification tasks and clustering algorithms. This usually means you have access to a researchers implementation of a new algorithm fairly quickly, long before it's in a commercial package. It also means you have to deal with a lot of 1-off code and deciding whether their function wants a row-vector or column-vector.

Python seems to be much more popular with those having a computer science background. There are far fewer machine learning algorithms available in Python. However, if you are going to design a large system, it's generally much more convenient to do in Python. There are Python interfaces to R as well.

Julia is new on the scene and attempts to solve the shortcomings of Python and R (insert xkcd comic here). Performance is good and has interfaces to many languages. I've used it a few times and it's maturing, but it's definitely risky doing any long term project in Julia.

Then there's Java. Weka is a popular machine learning package with a GUI and all of the algorithms available as jar files. Very consistent API and includes pre-processing tools. Weka also has a marketplace for new algorithms. However, many times you just have to write a 1-time script for data cleanup or to compare algorithms, and it's definitely not convenient to do in Java. I haven't seen many pure Java people doing this type of work in the wild. The final implementation may end up in Java, but the initial work seems to almost always be in R and Python.

Comment Re: The AI fanatics must be getting really despera (Score 2) 56

I think many have abandoned the term AI. There's too much history and it's misleading. Machine learning is more often used, but the phrase I think is most appropriate is statistical learning. It uses past data to predict future results (in the supervised world anyway). Different algorithms have different strengths and weaknesses.. Most serious researchers have also abandoned the idea of trying to strictly model the brain. We probably don't know enough about it to come even close, so in the meantime let's create algorithms that do useful things for us today.

Slashdot Top Deals

The IQ of the group is the lowest IQ of a member of the group divided by the number of people in the group.