Hacker News: aaggarwal

New comment by aaggarwal in "Uber, DoorDash and similar firms can’t defy the laws of capitalism after all"

aaggarwal — Wed, 10 Nov 2021 21:10:28 +0000

I see these articles every now and then over last 5-10 years, but as someone who has followed/worked around this industry in the past, I can tell you that there is enough demand, supply and capital for this industry and it is not going anywhere.

All of these companies are EBDITA profitable with decent margins and the net loss is them reinvesting in marketing, product and M&A (from their earnings reports). As the effects of covid wanes, their profits are going to go up very quickly which will be seen over the next year.

New comment by aaggarwal in "Rent control isn’t working in Sweden"

aaggarwal — Thu, 26 Aug 2021 17:24:12 +0000

City and state governments can offer temporary tax breaks for building new housing in certain zones to attract capital. This is already done in a way when government is trying to attract capital for new industries via SEZs.

New comment by aaggarwal in "Rent control isn’t working in Sweden"

aaggarwal — Thu, 26 Aug 2021 17:22:14 +0000

One problem with high tax based on capital gains at sale is that it needs the sale to happen and you start running into issues where people never sell and houses will get inherited. This still doesn’t tackle the problem head on.

New comment by aaggarwal in "Rent control isn’t working in Sweden"

aaggarwal — Thu, 26 Aug 2021 17:02:14 +0000

Most housing discussions seems to avoid one of the main reason driving the prices up, their treatment as an investment vehicle.

Residential housing should not be used as an investment vehicle. Period. Its primary purpose is to allow families to own and live in the house. Mandating that can eliminate the rich landlords/foreign investors buying houses just for investment, stabilize the demand and drive down the prices.

One way to mandate that could be progressive property tax based on how many properties you already own with a max cap on number of allowed residential properties one can own. This might decrease incentive to build new housing for developers, but can be solved by temporary tax breaks for building new housing.

Am I missing some second order effects?

A Principled Approach to Safety in Self-Driving – Uber ATG Safety Report

aaggarwal — Fri, 02 Nov 2018 20:19:48 +0000

Article URL: https://medium.com/@UberATG/a-principled-approach-to-safety-30dd0386a97c

Comments URL: https://news.ycombinator.com/item?id=18366381

Points: 2

# Comments: 0

A New Framework for Flexible and Reproducible Reinforcement Learning Research

aaggarwal — Wed, 29 Aug 2018 13:30:42 +0000

Article URL: https://ai.googleblog.com/2018/08/introducing-new-framework-for-flexible.html

Comments URL: https://news.ycombinator.com/item?id=17868053

Points: 5

# Comments: 0

My Fast and Furious Journey to Getting a Green Card

aaggarwal — Wed, 15 Aug 2018 11:30:08 +0000

Article URL: https://medium.com/@stellaguan/my-fast-and-furious-journey-to-getting-a-green-card-51dd0a9e4511

Comments URL: https://news.ycombinator.com/item?id=17765231

Points: 1

# Comments: 0

Uber CEO says Uber and Alphabet are discussing working together on self-driving

aaggarwal — Thu, 31 May 2018 09:12:25 +0000

Article URL: https://www.recode.net/2018/5/31/17390030/uber-ceo-dara-khosrowshahi-code-conference-interview

Comments URL: https://news.ycombinator.com/item?id=17195005

Points: 81

# Comments: 33

Show HN: MNIST-Sequence – Generate dataset for sequences of handwritten digits

aaggarwal — Mon, 13 Nov 2017 22:37:01 +0000

Article URL: https://github.com/ankitaggarwal011/MNIST-Sequence

Comments URL: https://news.ycombinator.com/item?id=15690943

Points: 5

# Comments: 0

New comment by aaggarwal in "A Competitive Programmer's Handbook"

aaggarwal — Sat, 15 Apr 2017 06:07:44 +0000

The author originally released the book here (http://codeforces.com/blog/entry/50728).

A Competitive Programmer's Handbook

aaggarwal — Fri, 14 Apr 2017 17:41:55 +0000

Article URL: https://cses.fi/book.html

Comments URL: https://news.ycombinator.com/item?id=14115826

Points: 1117

# Comments: 157

The future of the open internet – and our way of life – is in your hands

aaggarwal — Tue, 21 Mar 2017 10:31:07 +0000

Article URL: https://medium.freecodecamp.com/inside-the-invisible-war-for-the-open-internet-dd31a29a3f08#.8ffp3sc11

Comments URL: https://news.ycombinator.com/item?id=13920998

Points: 1

# Comments: 0

New comment by aaggarwal in "Show HN: Academic Bio – A simple static site generator for academic webpages"

aaggarwal — Wed, 02 Nov 2016 06:13:20 +0000

Hi James, this is pretty neat. I think this website could be a big hit among all the graduate applicants since they all are looking for creating a personal academic website at the time of applications.

Is there an option to import a latex resume to create the website? That would be handy.

I just want to thank you for ShareLatex, I use it for all my academic documents, it's pretty awesome.

New comment by aaggarwal in "Show HN: A flask app to make dashboards, easily"

aaggarwal — Sat, 10 Sep 2016 05:48:55 +0000

Yes, I used Freeboard (https://github.com/Freeboard/freeboard) couple of years ago for an IoT store. It's highly customizable and has fantastic UI. It can be used with minimal effort. Highly recommended.

The OP's project is also pretty cool.

New comment by aaggarwal in "PyCNN: Cellular Neural Networks Image Processing Python Library"

aaggarwal — Sun, 21 Aug 2016 07:56:31 +0000

For the people interested in reading about different template configuration and their results, this is a great resource:

[1] http://cnn-technology.itk.ppke.hu/Template_library_v3.1.pdf

[2] http://cnn-technology.itk.ppke.hu/Template_library_v4.0alpha...

New comment by aaggarwal in "PyCNN: Cellular Neural Networks Image Processing Python Library"

aaggarwal — Sun, 21 Aug 2016 07:47:09 +0000

Nice catch! It is indeed closely related to cellular automata. CNN processors could be thought of as a hybrid between ANN (artificial neural networks) and CA (continuous/cellular automata) [1]. Like neural networks, they are large-scale nonlinear analog circuits that process signals in real time. Like cellular automata, they consist of a massive aggregate of regularly spaced circuit clones, called cells, which communicate with each other directly only through their nearest neighbors [2].

The topology and dynamics of CNN processors closely resembles that of CA. Like most CNN processors, CA consists of a fixed-number of identical processors that are spatially discrete and topologically uniform. The difference is that most CNN processors are continuous-valued whereas CA have discrete-values [1].

You can also read this: http://www.nims.go.jp/nanophys6/Anirban%20Bandyopadhyay/site...

[1] https://en.wikipedia.org/wiki/Cellular_neural_network#Relate...

[2] http://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=7600

New comment by aaggarwal in "PyCNN: Cellular Neural Networks Image Processing Python Library"

aaggarwal — Sun, 21 Aug 2016 06:59:19 +0000

Yes, this is one of the applications of these networks.

New comment by aaggarwal in "PyCNN: Cellular Neural Networks Image Processing Python Library"

aaggarwal — Sat, 20 Aug 2016 18:14:20 +0000

Actually, it's more related to the original naming term that was adopted for cellular neural networks (CNN). Convolutional neural networks were originally referred to as ConvNets (CNN term is also used for them as well).

IMHO, I don't think this naming is misleading since CNN has been originally used to refer cellular neural networks.

New comment by aaggarwal in "PyCNN: Cellular Neural Networks Image Processing Python Library"

aaggarwal — Sat, 20 Aug 2016 18:08:44 +0000

Please take a look at the image processing specific applications [1] and advantages of the cellular neural networks.

[1] https://en.wikipedia.org/wiki/Cellular_neural_network#Applic...

New comment by aaggarwal in "PyCNN: Cellular Neural Networks Image Processing Python Library"

aaggarwal — Sat, 20 Aug 2016 18:03:15 +0000

Hi, author here. Even though I was on HN, I didn't realize that this is posted over here :) I'm glad about all the feedback. Thank you.

> I'm reading up on Cellular Neural Networks and it seems like they are simple convolutions, of which you can specify the kernel.

Actually its more than that, simply put, cellular neural networks are a parallel computing paradigm similar to neural networks, with the difference that communication is allowed between neighboring units only [1].

> In fact, this library is just calling scipy.signal.convolve2d() with different kernels.

The part you're referring to performs the convolution between the kernel function and the feedback template to get the result of the feedback loop. Please note the kernel function is sigmoidal or its approximation and remains unchanged.

It will be easier to understand if you'll visualize it as a control system as shown in [2] with a feedback template and a control template. These templates (coefficients) are configurable and produce different results for different configurations.

One of the applications of these networks is image processing as stated in [3] "CNN processors were designed to perform image processing; specifically, the original application of CNN processors was to perform real-time ultra-high frame-rate (>10,000 frame/s) processing unachievable by digital processors needed for applications like particle detection in jet engine fluids and spark-plug detection.".

[1] https://en.wikipedia.org/wiki/Cellular_neural_network

[2] http://www.isiweb.ee.ethz.ch/haenggi/CNN_web/CNN_figures/blo...

[3] https://en.wikipedia.org/wiki/Cellular_neural_network#Applic...