Do all the things like ++ or -- rants, post your own rants, comment on others' rants and build your customized dev avatarSign Up
From the creators of devRant, Pipeless lets you power real-time personalized recommendations and activity feeds using a simple APILearn More
Search - "deep neural networks"
Smart India Hackathon: Horrible experience
Background:- Our task was to do load forecasting for a given area. Hourly energy consumption data for past 5 years was given to us.
One government official asks the following questions:-
1. Why are you using deep learning for the project? Why are you not doing data analysis?
2. Which neural network "algorithm" you are using? He wanted to ask which model we are using, but he didn't have a single clue about Neural Networks.
3. Why are you using libraries? Why not your own code?
Here comes the biggest one,
4. Why haven't you developed your own "algorithm" (again, he meant model)? All you have done is used sone library. Where is "novelty" in your project?
I just want to say that if you don't know anything about ML/AI, then don't comment anything about it. And worst thing was, he was not ready to accept the fact that for capturing temporal dependencies where underlying probability distribution ia unknown, deep learning performs much better than traditional data analysis techniques.
After hearing his first question, second one was not a surprise for us. We were expecting something like that. For a few moments, we were speechless. Then one of us started by showing neural network architecture. But after some time, he rudely repeated the same question, "where is the algorithm". We told him every fucking thing used in the project, ranging from RMSprop optimizer to Backpropagation through time algorithm to mean squared loss error function.
Then very calmly, he asked third question, why are you using libraries? That moron wanted us to write a whole fucking optimized library. We were speechless at this question. Finally, one of us told him the "obvious" answer. We were completely demotivated. But it didnt end here. The real question was waiting. At the end, after listening to all of us, he dropped the final bomb, WHY HAVE YOU USED A NEURAL NETWORK "ALGORITHM" WHICH HAS ALREADY BEEN IMPLEMENTED? WHY DIDN'T YOU MAKE YOU OWN "ALGORITHM"? We again stated the obvious answer that it takes atleast an year or two of continuous hardwork to develop a state of art algorithm, that too when gou build it on top of some existing "algorithm". After listening to this, he left. His final response was "Try to make a new "algorithm"".
Needless to say, we were completely demotivated after this evaluation. We all had worked too hard for this. And we had ability to explain each and every part of the project intuitively and mathematically, but he was not even ready to listen.
Now, all of us are sitting aimlessly, waiting for Hackathon to end.😢😢😢😢😢25
(Warning: kinda long && somewhat of a political rant)
Every time I tell someone I work with AI, the first thing to come out of their mouth is "oh but AI is going to take over the world!"
It was only somewhat recently that it started being able to recognize what was in a picture from over 3 million images, and that too it's not that great at. Honestly people always say "AI is just if-else" ironically, but it isn't really that far from the truth, we just multiply an input by weights and check the output.
It isn't some magical sauce, it's not being born and then exploring a problem, it's just glorified-probability prediction. Even in "unsupervised" learning, the domain set is provided; in "reinforcement learning" which has gotten super popular lately we just have the computer decide which policy is optimal and apply that to an environment. It's a glorified decision tree (and technically tree models like XGBoost outperform neural networks and deep learning on a large number of problems) and it isn't going to "decide" to take over the planet.
Honestly all of this is just born out of Elon Musk fans who take his word as truth and have been led to believe that AI is going to take over the world. There are a billion reasons why it can't! And to top it off this takes away a lot of public attention from VERY concerning ethical issues with AI.
Am I the only one who saw Google Duplex being unveiled and immediately thought "fraud"? Forget phone scammers, if you trained duplex on the mannerisms of, for example, a famous politician's voice, you could impersonate them in an audio clip (or even video clip with deepfakes). Or for example the widespread use of object detection and facial recognition in surveillance systems deployed by DoD. Or the use of AI combined with location tracking and browsing analytics for targeted marketing.
The list of ethics breaches are endless, and I find it super suspicious that those profiting the most off of unethical AI are all too eager to shift public concern to some science fiction Terminator style takeover that, if ever possible, would be a long way out and is not any sort of a priority issue right now.11
Professor asks me to do research on deep complex neural networks, as in neural networks that perform on complex numbers.
Meanwhile me: "Google, what are complex numbers?"24
"Programming today is a race between software engineers striving to build bigger and better idiot-proof programs, and the universe trying to build bigger and better idiots. So far the universe is winning." ~~ Rick Cook
This guy single handedly explained GANs back in 90s and nobody noticed
Pro tip: always make sure your methods return the correct variable.
I’m currently working with deep neural networks using tensorflow. I needed to generate some test data and wrote a program to create it. I had two text files which each consisted of approximately 5000 lines of text.
I wrote a method that should sort out some words, and make my final data shorter. When I executed the program first time on our server, it spent about 25 minutes, then crashed due to MemoryError (which in Python means that the server didn’t have enough ram). That seemed quite weird since I only had about 10k lines of text, and I even sorted out a bunch of it, and the server has 128gb ram, and nothing’s using it.
Apparently I returned the wrong variable. That meant that my program tried to save 750 quadrillion lines of text rather than just a few thousand.
Always make sure to return the correct variables!1
(disclaimer: this might sound like a 13 year old guy just coming out from a theater after watching matrix but in some ways, its not )
Why the fuck should i feel discouraged from getting into ml/ai by all these smart ass people continuously taunting that "yeah, you might get into ml/ai but you won't get successful if you have a bad maths"
Ok, 1) i totally agree with these guys. Checked some pages and everywhere itse regression ,linear something, nlp's and neural networks, which even by the name sounds full of maths. BUT here is a thing:
1) All i can think of this as an ocean: just like web development is so vast, android is so vast, i can assume this to be so vast too.
BUT I WANT TO APPLY IT, NOT MAKE IT! why? Because that's what a beginner would do. It's data "sciences" and ppl who are deep into it will be called data "scientist" , a fuckin doctorate profession!
And toda i see it at so manypaces : from alexa playing song to google searches , youtube recommendation, hell even coffe machines are getting smarter!
I like these things and want to apply them as a developer to my apps and websites. But tell me, do everyone making a scanner or search engines learns regression algorithms and lambda calculas?
I love automation. So much that if given a chance, i would make robot to fuckin suck me off! From smart searches , self driving cars, map routes to latest apps with awesome pattern recognitions, i love them all. What i want to do is to look at some codes, tweak them for my usage and make something extraordinary and automated machine learning from ussr's interaction. I don't think my interest to learn applications of a technology and not the technology itself should be considered wrong because both are a carreer in their own! Learning ml/ai vs learning their applications is like a learning physics vs learning furniture designing: one being a part of other but completely existential on its own .
Thus the question comes what should i do? I got attracted by ML's achievements and fireworks but every ml course wants me to be the cracker maker! I want to get into data sciences bcz of its achievements ; and i want to replicate them again nd again until get termed as a professional nd if i feel the urge, maybe re visit my collage books and read maths and get into nlp designing (or whatever)
Where to get knowledge of this "life automation technologies" / data sciences (if they both are rea equal) and knowledge of such "ml toolkits" , if its really possible to be into ml without maths?8
Does gradient descent in artificial neural networks apply the most changes closest to the input layer?6
Machine Learning and Deep Neural Networks in particular. More job offers to pick from in upcoming years for me :P
"Deep is in. We want people to go deep. Deep neural networks … as opposed to shallow neural networks"
So I was asked by a client to make an app similar to prisma(not exactly that but let's say a caricature app) and I knew I have to research a lot.
Now I have been loyal to PHP for over 5 years so I first tried with GD and imagick but the results were not very good, so I thought let's try opencv. I didn’t wanna make any compromises so I didn't go the bridging way, I worked on native python even though I am a newbie in it. I was fairly impressed with the cartoonizing results but others weren't. Soon I got to know that this would take much more than simple filter combinations or matrix manipulations.
I read about prisma and got to know it uses deep neural networks for the same.
Now, in the five years I have learnt almost all the things a run-of-the-mill "Full stack Web Developer" should know.
I have a fair knowledge of PHP, many of its frameworks, many js frameworks(obviously jquery), I have a very good understanding of CSS and its models, I have worked on some cool algos and found solutions to many problems but I haven't gotten to stage where I can implement neural networks/machine learning in my projects.
It just scares me.
A little back story: I have been the CTO of a small scale company for about 1.5 years now.
So all this got me to asking myself should I just step down from the post to a position where I can learn more skills. Managing takes a lot more time where I can't learn a lot. Sure I learnt some other important things but not as much tech knowledge as I would have in a more basic position.
I know not many of you must have read this far, but if you did what do you think I should do? Really depressed at the moment.5
There are people who develop Neural Networks/Deep Learning Models/AI based Softwares.
Does anybody know what do we call them? Is it okay to call all of them Machine Learning Engineer/AI researcher/AI engineer?
If I'm looking for someone who can make AI based program for me. Whom should I be looking for on freelancer or LinkedIn?1