Following Ilyas Khan

Followed Ilyas Khan (LinkedIn)
Ilyas Khan Co-Founder and CEO at Cambridge Quantum Computing

Dear god, I wish Ilyas had a traditional blog with a true feed, but I’m willing to put up with the inconvenience of manually looking him up from time to time to see what he’s writing about quantum mechanics, quantum computing, category theory, and other areas of math.

Reply to A (very) gentle comment on Algebraic Geometry for the faint-hearted | Ilyas Khan

Replied to A (very) gentle comment on Algebraic Geometry for the faint-hearted by Ilyas KhanIlyas Khan (LinkedIn)
This short article is the result of various conversations over the course of the past year or so that arose on the back of two articles/blog pieces that I have previously written about Category Theory (here and here). One of my objectives with such articles, whether they be on aspects of quantum computing or about aspects of maths, is to try and de-mystify as much of the associated jargon as possible, and bring some of the stunning beauty and wonder of the subject to as wide an audience as possible. Whilst it is clearly not possible to become an expert overnight, and it is certainly not my objective to try and provide more than an introduction (hopefully stimulating further research and study), I remain convinced that with a little effort, non-specialists and even self confessed math-phobes can grasp some of the core concepts. In the case of my articles on Category Theory, I felt that even if I could generate one small gasp of excited comprehension where there was previously only confusion, then the articles were worth writing.

I just finished a course on Algebraic Geometry through UCLA Extension, which was geared toward non-traditional math students and professionals, and wish I had known about Smith’s textbook when I’d started. I did spend some time with Cox, Little, and O’Shea’s Ideals, Varieties, and Algorithms which is a pretty good introduction to the area, but written a bit more for computer scientists and engineers in mind rather than the pure mathematician, which might recommend it more toward your audience here as well. It’s certainly more accessible than Hartshorne for the faint-of-heart.

I’ve enjoyed your prior articles on category theory which have spurred me to delve deeper into the area. For others who are interested, I thought I’d also mention that physicist and information theorist John Carlos Baez at UCR has recently started an applied category theory online course which I suspect is a bit more accessible than most of the higher graduate level texts and courses currently out. For more details, I’d suggest starting here: https://johncarlosbaez.wordpress.com/2018/03/26/seven-sketches-in-compositionality/

Syndicated copies to:

🔖 Special Issue : Information Dynamics in Brain and Physiological Networks | Entropy

Bookmarked Special Issue "Information Dynamics in Brain and Physiological Networks" (mdpi.com)

A special issue of Entropy (ISSN 1099-4300). This special issue belongs to the section "Information Theory".

Deadline for manuscript submissions: 30 December 2018

It is, nowadays, widely acknowledged that the brain and several other organ systems, including the cardiovascular, respiratory, and muscular systems, among others, exhibit complex dynamic behaviors that result from the combined effects of multiple regulatory mechanisms, coupling effects and feedback interactions, acting in both space and time.

The field of information theory is becoming more and more relevant for the theoretical description and quantitative assessment of the dynamics of the brain and physiological networks, defining concepts, such as those of information generation, storage, transfer, and modification. These concepts are quantified by several information measures (e.g., approximate entropy, conditional entropy, multiscale entropy, transfer entropy, redundancy and synergy, and many others), which are being increasingly used to investigate how physiological dynamics arise from the activity and connectivity of different structural units, and evolve across a variety of physiological states and pathological conditions.

This Special Issue focuses on blending theoretical developments in the new emerging field of information dynamics with innovative applications targeted to the analysis of complex brain and physiological networks in health and disease. To favor this multidisciplinary view, contributions are welcome from different fields, ranging from mathematics and physics to biomedical engineering, neuroscience, and physiology.

Prof. Dr. Luca Faes
Prof. Dr. Alberto Porta
Prof. Dr. Sebastiano Stramaglia
Guest Editors
Syndicated copies to:

👓 The Scientific Paper Is Obsolete | The Atlantic

Read The Scientific Paper Is Obsolete by James Somers (The Atlantic)
The scientific paper—the actual form of it—was one of the enabling inventions of modernity. Before it was developed in the 1600s, results were communicated privately in letters, ephemerally in lectures, or all at once in books. There was no public forum for incremental advances. By making room for reports of single experiments or minor technical advances, journals made the chaos of science accretive. Scientists from that point forward became like the social insects: They made their progress steadily, as a buzzing mass.

The earliest papers were in some ways more readable than papers are today. They were less specialized, more direct, shorter, and far less formal. Calculus had only just been invented. Entire data sets could fit in a table on a single page. What little “computation” contributed to the results was done by hand and could be verified in the same way.

Not quite the cutting edge stuff I would have liked, but generally an interesting overview of relatively new technology and UI set ups like Mathematica and Jupyter.

Syndicated copies to:

🔖 PaperBadger

Bookmarked PaperBadger by Mozilla Science (GitHub)
Issuing badges to credit authors for their work on academic papers https://badges.mozillascience.org/

Exploring the use of digital badges for crediting contributors to scholarly papers for their work

As the research environment becomes more digital, we want to test how we can use this medium to help bring transparency and credit for individuals in the publication process.

This work is a collaboration with publishers BioMed Central (BMC), Ubiquity Press (UP) and the Public Library of Science (PLoS); the biomedical research foundation, The Wellcome Trust; the software and technology firm Digital Science; the registry of unique researcher identifiers, ORCID; and the Mozilla Science Lab.

h/t to Greg McVerry via https://chat.indieweb.org/dev/2018-04-04#t1522869725219200

🔖 [1803.08823] A high-bias, low-variance introduction to Machine Learning for physicists | arXiv

Bookmarked A high-bias, low-variance introduction to Machine Learning for physicists by Pankaj Mehta, Marin Bukov, Ching-Hao Wang, Alexandre G.R. Day, Clint Richardson, Charles K. Fisher, David J. Schwab (arxiv.org)
Machine Learning (ML) is one of the most exciting and dynamic areas of modern research and application. The purpose of this review is to provide an introduction to the core concepts and tools of machine learning in a manner easily understood and intuitive to physicists. The review begins by covering fundamental concepts in ML and modern statistics such as the bias-variance tradeoff, overfitting, regularization, and generalization before moving on to more advanced topics in both supervised and unsupervised learning. Topics covered in the review include ensemble models, deep learning and neural networks, clustering and data visualization, energy-based models (including MaxEnt models and Restricted Boltzmann Machines), and variational methods. Throughout, we emphasize the many natural connections between ML and statistical physics. A notable aspect of the review is the use of Python notebooks to introduce modern ML/statistical packages to readers using physics-inspired datasets (the Ising Model and Monte-Carlo simulations of supersymmetric decays of proton-proton collisions). We conclude with an extended outlook discussing possible uses of machine learning for furthering our understanding of the physical world as well as open problems in ML where physicists maybe able to contribute. (Notebooks are available at this https URL )
Syndicated copies to:

👓 Mutating DNA caught on film | Science | AAAS

Read Mutating DNA caught on film by Elizabeth Pennisi (Science | AAAS)
Study in bacteria shows how regularly DNA changes and how few of those changes are deadly

This is a rather cool little experiment.

h/t to @moorejh via Twitter:

Bookmarked on March 16, 2018 at 12:15PM

Syndicated copies to:

👓 Living Bits: Information and the Origin of Life | PBS

Read Living Bits: Information and the Origin of Life by Christoph Adami (pbs.org)
What is life? When Erwin Schrödinger posed this question in 1944, in a book of the same name, he was 57 years old. He had won the Nobel in Physics eleven years earlier, and was arguably past his glory days. Indeed, at that time he was working mostly on his ill-fated “Unitary Field Theory.” By all accounts, the publication of “What is Life?”—venturing far outside of a theoretical physicist’s field of expertise—raised many eyebrows. How presumptuous for a physicist to take on one of the deepest questions in biology! But Schrödinger argued that science should not be compartmentalized: “Some of us should venture to embark on a synthesis of facts and theories, albeit with second-hand and incomplete knowledge of some of them—and at the risk of making fools of ourselves.” Schrödinger’s “What is Life” has been extraordinarily influential, in one part because he was one of the first who dared to ask the question seriously, and in another because it was the book that was read by a good number of physicists—famously both Francis Crick and James Watson independently, but also many a member of the “Phage group,” a group of scientists that started the field of bacterial genetics—and steered them to new careers in biology. The book is perhaps less famous for the answers Schrödinger suggested, as almost all of them have turned out to be wrong.

Highlights, Quotes, & Marginalia

our existence can succinctly be described as “information that can replicate itself,” the immediate follow-up question is, “Where did this information come from?”

from an information perspective, only the first step in life is difficult. The rest is just a matter of time.

Through decades of work by legions of scientists, we now know that the process of Darwinian evolution tends to lead to an increase in the information coded in genes. That this must happen on average is not difficult to see. Imagine I start out with a genome encoding n bits of information. In an evolutionary process, mutations occur on the many representatives of this information in a population. The mutations can change the amount of information, or they can leave the information unchanged. If the information changes, it can increase or decrease. But very different fates befall those two different changes. The mutation that caused a decrease in information will generally lead to lower fitness, as the information stored in our genes is used to build the organism and survive. If you know less than your competitors about how to do this, you are unlikely to thrive as well as they do. If, on the other hand, you mutate towards more information—meaning better prediction—you are likely to use that information to have an edge in survival.

There are some plants with huge amounts of DNA compared to their “peers”–perhaps these would be interesting test cases for potential experimentation of this?

Syndicated copies to:

🔖 Efficient Algorithms for Searching the Minimum Information Partition in Integrated Information Theory

Bookmarked Efficient Algorithms for Searching the Minimum Information Partition in Integrated Information Theory by Jun Kitazono, Ryota Kanai, Masafumi Oizumi (MDPI)
The ability to integrate information in the brain is considered to be an essential property for cognition and consciousness. Integrated Information Theory (IIT) hypothesizes that the amount of integrated information ( Φ ) in the brain is related to the level of consciousness. IIT proposes that, to quantify information integration in a system as a whole, integrated information should be measured across the partition of the system at which information loss caused by partitioning is minimized, called the Minimum Information Partition (MIP). The computational cost for exhaustively searching for the MIP grows exponentially with system size, making it difficult to apply IIT to real neural data. It has been previously shown that, if a measure of Φ satisfies a mathematical property, submodularity, the MIP can be found in a polynomial order by an optimization algorithm. However, although the first version of Φ is submodular, the later versions are not. In this study, we empirically explore to what extent the algorithm can be applied to the non-submodular measures of Φ by evaluating the accuracy of the algorithm in simulated data and real neural data. We find that the algorithm identifies the MIP in a nearly perfect manner even for the non-submodular measures. Our results show that the algorithm allows us to measure Φ in large systems within a practical amount of time.

h/t Christoph Adami, Erik Hoel, and @kanair

👓 Stephen Hawking, Who Examined the Universe and Explained Black Holes, Dies at 76 | The New York Times

Read Stephen Hawking, Who Examined the Universe and Explained Black Holes, Dies at 76 by Dennis Overbye (nytimes.com)
A physicist and best-selling author, Dr. Hawking did not allow his physical limitations to hinder his quest to answer “the big question: Where did the universe come from?”

Some sad news after getting back from Algebraic Geometry class tonight. RIP Stephen Hawking.

Syndicated copies to:

The Physics of Life: Summer School | Center for the Physics of Biological Function

Bookmarked The Physics of Life: Summer School | Center for the Physics of Biological Function (biophysics.princeton.edu)
A summer school for advanced undergraduates June 11-22, 2018 @ Princeton University What would it mean to have a physicist’s understanding of life? How do DYNAMICS and the EMERGENCE of ORDER affect biological function? How do organisms process INFORMATION, LEARN, ADAPT, and EVOLVE? See how physics problems emerge from thinking about developing embryos, communicating bacteria, dynamic neural networks, animal behaviors, evolution, and more. Learn how ideas and methods from statistical physics, simulation and data analysis, optics and microscopy connect to diverse biological phenomena. Explore these questions, tools, and concepts in an intense two weeks of lectures, seminars, hands-on exercises, and projects.
Syndicated copies to:

Organizing my research related reading

There’s so much great material out there to read and not nearly enough time. The question becomes: “How to best organize it all, so you can read even more?”

I just came across a tweet from Michael Nielsen about the topic, which is far deeper than even a few tweets could do justice to, so I thought I’d sketch out a few basic ideas about how I’ve been approaching it over the last decade or so. Ideally I’d like to circle back around to this and better document more of the individual aspects or maybe even make a short video, but for now this will hopefully suffice to add to the conversation Michael has started.

Keep in mind that this is an evolving system which I still haven’t completely perfected (and may never), but to a great extent it works relatively well and I still easily have the ability to modify and improve it.

Overall Structure

The first piece of the overarching puzzle is to have a general structure for finding, collecting, triaging, and then processing all of the data. I’ve essentially built a simple funnel system for collecting all the basic data in the quickest manner possible. With the basics down, I can later skim through various portions to pick out the things I think are the most valuable and move them along to the next step. Ultimately I end up reading the best pieces on which I make copious notes and highlights. I’m still slowly trying to perfect the system for best keeping all this additional data as well.

Since I’ve seen so many apps and websites come and go over the years and lost lots of data to them, I far prefer to use my own personal website for doing a lot of the basic collection, particularly for online material. Toward this end, I use a variety of web services, RSS feeds, and bookmarklets to quickly accumulate the important pieces into my personal website which I use like a modern day commonplace book.

Collecting

In general, I’ve been using the Inoreader feed reader to track a large variety of RSS feeds from various clearinghouse sources (including things like ProQuest custom searches) down to individual researcher’s blogs as a means of quickly pulling in large amounts of research material. It’s one of the more flexible readers out there with a huge number of useful features including the ability to subscribe to OPML files, which many readers don’t support.

As a simple example arXiv.org has an RSS feed for the topic of “information theory” at http://arxiv.org/rss/math.IT which I subscribe to. I can quickly browse through the feed and based on titles and/or abstracts, I can quickly “star” the items I find most interesting within the reader. I have a custom recipe set up for the IFTTT.com service that pulls in all these starred articles and creates new posts for them on my WordPress blog. To these posts I can add a variety of metadata including top level categories and lower level tags in addition to other additional metadata I’m interested in.

I also have similar incoming funnel entry points via many other web services as well. So on platforms like Twitter, I also have similar workflows that allow me to use services like IFTTT.com or Zapier to push the URLs easily to my website. I can quickly “like” a tweet and a background process will suck that tweet and any URLs within it into my system for future processing. This type of workflow extends to a variety of sites where I might consume potential material I want to read and process. (Think academic social services like Mendeley, Academia.com, Diigo, or even less academic ones like Twitter, LinkedIn, etc.) Many of these services often have storage ability and also have simple browser bookmarklets that allow me to add material to them. So with a quick click, it’s saved to the service and then automatically ported into my website almost without friction.

My WordPress-based site uses the Post Kinds Plugin which takes incoming website URLs and does a very solid job of parsing those pages to extract much of the primary metadata I’d like to have without requiring a lot of work. For well structured web pages, it’ll pull in the page title, authors, date published, date updated, synopsis of the page, categories and tags, and other bits of data automatically. All these fields are also editable and searchable. Further, the plugin allows me to configure simple browser bookmarklets so that with a simple click on a web page, I can pull its URL and associated metadata into my website almost instantaneously. I can then add a note or two about what made me interested in the piece and save it for later.

Note here, that I’m usually more interested in saving material for later as quickly as I possibly can. In this part of the process, I’m rarely ever interested in reading anything immediately. I’m most interested in finding it, collecting it for later, and moving on to the next thing. This is also highly useful for things I find during my busy day that I can’t immediately find time for at the moment.

As an example, here’s a book I’ve bookmarked to read simply by clicking “like” on a tweet I cam across late last year. You’ll notice at the bottom of the post, I’ve optionally syndicated copies of the post to other platforms to “spread the wealth” as it were. Perhaps others following me via other means may see it and find it useful as well?

Triaging

At regular intervals during the week I’ll sit down for an hour or two to triage all the papers and material I’ve been sucking into my website. This typically involves reading through lots of abstracts in a bit more detail to better figure out what I want to read now and what I’d like to read at a later date. I can delete out the irrelevant material if I choose, or I can add follow up dates to custom fields for later reminders.

Slowly but surely I’m funneling down a tremendous amount of potential material into a smaller, more manageable amount that I’m truly interested in reading on a more in-depth basis.

Document storage

Calibre with GoodReads sync

Even for things I’ve winnowed down, there is still a relatively large amount of material, much of it I’ll want to save and personally archive. For a lot of this function I rely on the free multi-platform desktop application Calibre. It’s essentially an iTunes-like interface, but it’s built specifically for e-books and other documents.

Within it I maintain a small handful of libraries. One for personal e-books, one for research related textbooks/e-books, and another for journal articles. It has a very solid interface and is extremely flexible in terms of configuration and customization. You can create a large number of custom libraries and create your own searchable and sort-able fields with a huge variety of metadata. It often does a reasonable job of importing e-books, .pdf files, and other digital media and parsing out their meta data which prevents one from needing to do some of that work manually. With some well maintained metadata, one can very quickly search and sort a huge amount of documents as well as quickly prioritize them for action. Additionally, the system does a pretty solid job of converting files from one format to another, so that things like converting an .epub file into a .mobi format for Kindle are automatic.

Calibre stores the physical documents either in local computer storage, or even better, in the cloud using any of a variety of services including Dropbox, OneDrive, etc. so that one can keep one’s documents in the cloud and view them from a variety of locations (home, work, travel, tablet, etc.)

I’ve been a very heavy user of GoodReads.com for years to bookmark and organize my physical and e-book library and anti-libraries. Calibre has an exceptional plugin for GoodReads that syncs data across the two. This (and a few other plugins) are exceptionally good at pulling in missing metadata to minimize the amount that must be done via hand, which can be tedious.

Within Calibre I can manage my physical books, e-books, journal articles, and a huge variety of other document related forms and formats. I can also use it to further triage and order the things I intend to read and order them to the nth degree. My current Calibre libraries have over 10,000 documents in them including over 2,500 textbooks as well as records of most of my 1,000+ physical books. Calibre can also be used to add document data that one would like to ultimately acquire the actual documents, but currently don’t have access to.

BibTeX and reference management

In addition to everything else Calibre also has some well customized pieces for dovetailing all its metadata as a reference management system. It’ll allow one to export data in a variety of formats for document publishing and reference management including BibTex formats amongst many others.

Reading, Annotations, Highlights

Once I’ve winnowed down the material I’m interested in it’s time to start actually reading. I’ll often use Calibre to directly send my documents to my Kindle or other e-reading device, but one can also read them on one’s desktop with a variety of readers, or even from within Calibre itself. With a click or two, I can automatically email documents to my Kindle and Calibre will also auto-format them appropriately before doing so.

Typically I’ll send them to my Kindle which allows me a variety of easy methods for adding highlights and marginalia. Sometimes I’ll read .pdf files via desktop and use Adobe to add highlights and marginalia as well. When I’m done with a .pdf file, I’ll just resave it (with all the additions) back into my Calibre library.

Exporting highlights/marginalia to my website

For Kindle related documents, once I’m finished, I’ll use direct text file export or tools like clippings.io to export my highlights and marginalia for a particular text into simple HTML and import it into my website system along with all my other data. I’ve briefly written about some of this before, though I ought to better document it. All of this then becomes very easily searchable and sort-able for future potential use as well.

Here’s an example of some public notes, highlights, and other marginalia I’ve posted in the past.

Synthesis

Eventually, over time, I’ve built up a huge amount of research related data in my personal online commonplace book that is highly searchable and sortable! I also have the option to make these posts and pages public, private, or even password protected. I can create accounts on my site for collaborators to use and view private material that isn’t publicly available. I can also share posts via social media and use standards like webmention and tools like brid.gy so that comments and interactions with these pieces on platforms like Facebook, Twitter, Google+, and others is imported back to the relevant portions of my site as comments. (I’m doing it with this post, so feel free to try it out yourself by commenting on one of the syndicated copies.)

Now when I’m ready to begin writing something about what I’ve read, I’ve got all the relevant pieces, notes, and metadata in one centralized location on my website. Synthesis becomes much easier. I can even have open drafts of things as I’m reading and begin laying things out there directly if I choose. Because it’s all stored online, it’s imminently available from almost anywhere I can connect to the web. As an example, I used a few portions of this workflow to actually write this post.

Continued work

Naturally, not all of this is static and it continues to improve and evolve over time. In particular, I’m doing continued work on my personal website so that I’m able to own as much of the workflow and data there. Ideally I’d love to have all of the Calibre related piece on my website as well.

Earlier this week I even had conversations about creating new post types on my website related to things that I want to read to potentially better display and document them explicitly. When I can I try to document some of these pieces either here on my own website or on various places on the IndieWeb wiki. In fact, the IndieWeb for Education page might be a good place to start browsing for those interested.

One of the added benefits of having a lot of this data on my own website is that it not only serves as my research/data platform, but it also has the traditional ability to serve as a publishing and distribution platform!

Currently, I’m doing most of my research related work in private or draft form on the back end of my website, so it’s not always publicly available, though I often think I should make more of it public for the value of the aggregation nature it has as well as the benefit it might provide to improving scientific communication. Just think, if you were interested in some of the obscure topics I am and you could have a pre-curated RSS feed of all the things I’ve filtered through piped into your own system… now multiply this across hundreds of thousands of other scientists? Michael Nielsen posts some useful things to his Twitter feed and his website, but what I wouldn’t give to see far more of who and what he’s following, bookmarking, and actually reading? While many might find these minutiae tedious, I guarantee that people in his associated fields would find some serious value in it.

I’ve tried hundreds of other apps and tools over the years, but more often than not, they only cover a small fraction of the necessary moving pieces within a much larger moving apparatus that a working researcher and writer requires. This often means that one is often using dozens of specialized tools upon which there’s a huge duplication of data efforts. It also presumes these tools will be around for more than a few years and allow easy import/export of one’s hard fought for data and time invested in using them.

If you’re aware of something interesting in this space that might be useful, I’m happy to take a look at it. Even if I might not use the service itself, perhaps it’s got a piece of functionality that I can recreate into my own site and workflow somehow?

If you’d like help in building and fleshing out a system similar to the one I’ve outlined above, I’m happy to help do that too.

Related posts

Syndicated copies to:
Acquired A New Kind of Science by Stephen Wolfram (Wolfram Media)
Starting from a collection of simple computer experiments illustrated by striking computer graphics Stephen Wolfram shows in this landmark book how their unexpected results force a whole new way of looking at the operation of our universe. Wolfram uses his approach to tackle a remarkable array of fundamental problems in science, from the origins of apparent randomness in physical systems, to the development of complexity in biology, the ultimate scope and limitations of mathematics, the possibility of a truly fundamental theory of physics, the interplay between free will and determinism, and the character of intelligence in the universe.

Gifted to me by my friend Dave Snead who picked up a copy from the Wolfram booth earlier today at the APS Conference in downtown Los Angeles. Thanks Dave!

Syndicated copies to:

👓 Science’s Inference Problem: When Data Doesn’t Mean What We Think It Does | New York Times

Read Science’s Inference Problem: When Data Doesn’t Mean What We Think It Does by James Ryerson (nytimes.com)
Three new books on the challenge of drawing confident conclusions from an uncertain world.

Not sure how I missed this when it came out two weeks ago, but glad it popped up in my reader today.

This has some nice overview material for the general public on probability theory and science, but given the state of research, I’d even recommend this and some of the references to working scientists.

I remember bookmarking one of the texts back in November. This is a good reminder to circle back and read it.

Syndicated copies to: