My new year’s resolutions—2020 edition

Here is a list of my resolutions for the new year…

1. Data Science:

1.1 A set of brief notes:

Write a small set of brief notes (in LaTeX) on topics from machine learning. Focus only on the broad nature of the problem, the essential mathematical/modeling ideas, and only outlines of algorithms. In the first cut, ignore detailed explanations, comments, applications, or detailed comparisons among techniques suitable for the same task.

The goal this year is mostly to further increase the breadth of my reading, understanding, and implementation in code, not depth.

Level of writing: UG. However, only for myself. (Won’t explain everything; the notes may look uneven/partial/not very smoothly progressing.) Wouldn’t care for people concluding: “this is all this guy knows, huh?” Just plain ignore them, and focus on the breadth.

These initial notes could start in any form, but ideally, they should give just a framework. It could be augmented or made deeper later on. (Some parts of it, right in this (i.e. 2020) year.)

Review the progress every two months.

1.2 New ideas:

Definitely work on my possibly new ideas (algorithms/approach), one each from canoncial ML, and DL. I believe they sound sound.

If really new, consider publishing at a good international conference. (Too many conferences in this field in India. And, too many of them too costly too. Only ॠषितुल्य’s can afford them.)

Review the progress every four months.

2. Quantum Mechanics:

The series of posts on ontologies of physics has already been converted into a standalone LaTeX document.

Revise it. Take a printout. Mark the points to keep, to eliminate, to reorder, etc. Also make point-wise notes on a separate piece of paper. Effect the revision into the LaTeX draft. Go through the loop a few times.

Decide on the best way to present my solution to the QM measurement problem. At least 2/3 different ways are possible (if not more); try out them all first in brief rough drafts. Then collect the points together and write a standalone document on the solution I propose (assuming the context of the ontologies document).

Revise the Outline document posted at iMechanica.

Desirable in the new year, but not a resolution: Convert the revised Outline document (the last one) into a first draft for a journal paper.

No specific time-schedule for this one activity. Only a review sure to be taken every month-end.

3. Health etc.:

Follow this principle: Keeping on trying at keeping at it, is more important than being regular. (Else, in resolving to be regular, once the routine breaks, it tends to stay broken for the rest of the year too. So, be easy on yourself. Further change the goal itself. Accordingly:

Minimum: Go for walks (30 minutes) on at least six days each month (“1.5 days a week”).

Separately, do surya namaskars on at least two days a week (eight times a month). (Preferably, walking and surya namaskars should be done on different days. Adjust between the two suitably during the monsoons.)

Keep a register of the days the activities were actually done. Review every two months, and decide at such times about suitably increasing the frequency for the next two months’ period.

4. Blogging:

No separate blog on Data Science for now.

Post the notes on data science at my personal Web site [^], with entries here linking to the same. Consider keeping the actual documents at my Google drive (or GitHub) and giving only links from the personal Web site or this blog (because if a PDF having the same name is updated, WordPress changes the name).

No limit on the blogging related to QM, whether in terms of number of posts or their lengths. When it comes to knowledge-development, atrophy is unacceptable—it’s best left to others.

Leaving aside QM and DS posts/documents, limit all the rest of the blogging to about 3–5 posts per quarter; the lower the better. Review every quarter.

In the songs section, if an already run song happens to strike me again, re-run it. I began adding this section in 2010. So far, I tried not to repeat songs, relying mostly on memory. Succeeded almost always, except for one inadvertent re-run (I believe).

Or, not run any such thing at all!

5. उपनिषदे (Upanishads):

Go through a few of these, and select a suitable one. (Within the first three months.)

Write a first rough draft over the rest of the year.

No need to blog about it right away. (But take a review about this goal sometime around mid-year.)

6. Meditation, side-readings (on all topics—not just spirituality), etc.:

Satisfactory pace achieved already. No need to change it; certainly no need to make NYR’s about them as such.

What else…

Oh! Happy new year!


A song I like:

[The first time I heard this song was when I bought a CD of a compilation of songs, right within the first few months once I was in the SF Bay Area (aka “Silicon Valley”). This was in the first part of1998.

I had not heard this particular RD song before, and was thankful to the compiler (a Sardaar, I vaguely remember) for including it in the compiled CD.

There never was any question of thinking about whether I like this song or not, right from the first time I heard it. As they say, it was “right up my alley”. The only noticeable thing was a hurt sense of pride. I had thought I had heard every (good) song that RD had composed. (In the COEP hostels, had become something of a “goto” guy to check if some song had been composed by SD, RD, or Rajesh Roshan/Bappi Lahiri.)

On possible reason I didn’t know about this song was: There was shooting done for it, but the song was not finally included in the as released movie. In any case, I am sure I had never heard it before that moment in 1998. Even if it might have infrequently been played on the radio, I had missed it.

Almost two decades later (in 2017 or 2018), I ran into another version of this song—done right by the original team. It feels like a rehearsal version. This is the version that was released on YouTube by SaReGaMa. The “released” CD version (which I heard) seems to be the final one.

I am not too sure which one I like better. I tend to think that it is the rehearsal version which I like better.

One final bit. People say that this was based on Charlie Chaplin’s Limelight theme. I had heard the Limelight theme long time before 1998, and I had loved it too. But, somehow, when I heard this song, I never made the connection. I just nodded that it was RD’s song. … But, yeah, after being told, it does make sense.

Links: The first link on Google search (the same version as in the compiled CD version): [^]. The “rehearsal” version, as released by SaReGaMa [^]. One of the comments which alludes to the Limelight connection [^]. ]

(Hindi) “tum meri zindagi mein kuchh” (तुम मेरी जिंदगी मे कुछ इस तरह से आये…)
Music: R. D. Burman
Lyrics: Rajinder Kishen
Singer: Lata Mangeshkar, Kishor Kumar