index.xml

<?xml version="1.0" encoding="utf-8" standalone="yes" ?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>S.R.A on S.R.A</title>
    <link>https://arshishir.github.io/</link>
    <description>Recent content in S.R.A on S.R.A</description>
    <generator>Hugo -- gohugo.io</generator>
    <language>en-us</language>
    <copyright>&amp;copy; 2019</copyright>
    <lastBuildDate>Sun, 15 Oct 2017 00:00:00 -0400</lastBuildDate>
    <atom:link href="/" rel="self" type="application/rss+xml" />
    
    <item>
      <title>Thermodynamics, ESM, and NESM</title>
      <link>https://arshishir.github.io/post/basic/</link>
      <pubDate>Tue, 02 Apr 2019 00:00:00 -0400</pubDate>
      
      <guid>https://arshishir.github.io/post/basic/</guid>
      <description>&lt;p&gt;The picture below summarizes very basic idea behind Thermodynamics, Equilibrium Statistical Mechanics, and Nonequilibrium Statistical Mechanics.



&lt;figure&gt;

&lt;img src=&#34;idea.jpg&#34; /&gt;



&lt;figcaption data-pre=&#34;Figure &#34; data-post=&#34;:&#34; &gt;
  &lt;h4&gt;Basic Idea based on Elements of Nonequilibrium Statistical Mechanics book by V. Balakrishnan&lt;/h4&gt;
  
&lt;/figcaption&gt;

&lt;/figure&gt;&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>Exploration in ML</title>
      <link>https://arshishir.github.io/post/expo_in_ml/</link>
      <pubDate>Sun, 03 Feb 2019 00:00:00 -0500</pubDate>
      
      <guid>https://arshishir.github.io/post/expo_in_ml/</guid>
      <description>

&lt;p&gt;For the past year or so, I have been dabbling in machine learning. Last summer, I decided to spend some serious amount of time exploring ML. As a beginner, I got overwhelmed by the plethora of resources out there. Meanwhile, I convinced one of my friends to go on this exploration with me. We decided to go through lectures on &lt;em&gt;machine learning by Andrew Ng&lt;/em&gt;. The idea was to watch his old &lt;em&gt;Coursera&lt;/em&gt; lectures and meet once a week and discuss those ideas.&lt;/p&gt;

&lt;p&gt;Fortunately that summer, I also got an opportunity to mentor a high school student. He wanted to explore some computer science related topic. So, I decided to teach him machine learning. This decision forced me to pick a book and some sets of lecture notes. I choose &lt;em&gt;Pattern Recognition and Machine Learning (PRML)&lt;/em&gt; by Christopher Bishop and &lt;em&gt;Deep learning specialization by Andrew Ng&lt;/em&gt;. With the added responsibility of teaching, I strated going through first few chapters of &lt;em&gt;PRML&lt;/em&gt;. Based on &lt;em&gt;PRML&lt;/em&gt; and &lt;em&gt;Deep Learning specialization&lt;/em&gt;, I made lecture notes so that I could convey those ideas clearly. It was fun!&lt;/p&gt;

&lt;p&gt;In the mean that I came across some other very useful resources. Here’s a list of some of those resources:&lt;/p&gt;

&lt;h3 id=&#34;support-vector-machine&#34;&gt;Support Vector Machine&lt;/h3&gt;

&lt;ul&gt;
&lt;li&gt;&lt;p&gt;One of the most gentle and clear introductions to SVM by &lt;a href=&#34;https://www.youtube.com/watch?v=_PwhiWxHK8o&amp;amp;t&#34; target=&#34;_blank&#34;&gt;Patrick Winston&lt;/a&gt;&lt;/p&gt;&lt;/li&gt;

&lt;li&gt;&lt;p&gt;Interesting and very helpful SVM recitation session by &lt;a href=&#34;https://www.youtube.com/watch?v=ik7E7r2a1h8&#34; target=&#34;_blank&#34;&gt;Jessica Noss&lt;/a&gt;&lt;/p&gt;&lt;/li&gt;
&lt;/ul&gt;

&lt;h3 id=&#34;machine-learning-lectures-by-nando-freitas&#34;&gt;Machine Learning lectures by Nando Freitas&lt;/h3&gt;

&lt;p&gt;I haven’t gone through the whole lecture series. However, I watched quite a few lectures by him and they are some of the clearest &lt;a href=&#34;https://www.youtube.com/watch?v=w2OtwL5T1ow&amp;amp;list=PLE6Wd9FR--EdyJ5lbFl8UuGjecvVw66F6&#34; target=&#34;_blank&#34;&gt;introductions those topics on ML&lt;/a&gt; that I found on youtube.&lt;/p&gt;

&lt;h3 id=&#34;machine-learning-lectures-by-ali-ghodsi&#34;&gt;Machine Learning lectures by Ali Ghodsi&lt;/h3&gt;

&lt;p&gt;One of the hidden gems of youtube &lt;a href=&#34;https://www.youtube.com/channel/UCKJNzy_GuvX3SAg3ipaGa8A/playlists&#34; target=&#34;_blank&#34;&gt;Machine learning lecture series&lt;/a&gt;
. I have watched a couple of this lecture and they are brilliant. I would highly recommend his lecture on Decision Tree.&lt;/p&gt;

&lt;div class=&#34;alert alert-note&#34;&gt;
  &lt;p&gt;This is by no means a complete list. I went through many more resources. I will add those and the ones that I am currently going through soon.&lt;/p&gt;

&lt;/div&gt;

</description>
    </item>
    
    <item>
      <title>Cracking Captcha using Deep Learning</title>
      <link>https://arshishir.github.io/projects/cracking_captach/</link>
      <pubDate>Fri, 01 Feb 2019 00:00:00 -0500</pubDate>
      
      <guid>https://arshishir.github.io/projects/cracking_captach/</guid>
      <description>

&lt;h2 id=&#34;status&#34;&gt;Status&lt;/h2&gt;

&lt;p&gt;Experimental &amp;amp; Ongoing&lt;/p&gt;

&lt;h2 id=&#34;research&#34;&gt;Research&lt;/h2&gt;

&lt;p&gt;As an academician, we have been mostly exploring theoretical aspects of machine learning. This project grew out of our need to get our hands dirty. So, the idea is to explore deep learning by figuring out how to crack captcha. We will play with different DNN models. Here, the idea is not to come up with an original method but to use previously tested models. Hopefully, while doing so we learn how to code up the model in different platforms and develop a work flow for machine learning projects.&lt;/p&gt;

&lt;h2 id=&#34;tools&#34;&gt;Tools&lt;/h2&gt;

&lt;p&gt;For scientific computing, we have have been using $Mathematica$. So, firstly we will $Mathematica$’s inbuilt Neural Network functions and tools.&lt;/p&gt;

&lt;p&gt;We have used $Python$ for data generation. We would also like to use $Python$’s ML packages like: $PyTorch, Keras, Fastai$.&lt;/p&gt;

&lt;p&gt;Lastly, we would like to use $Julia$’s ML packages like: $Flux, KNet$.&lt;/p&gt;

&lt;h2 id=&#34;procedure&#34;&gt;Procedure&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;Using the following python code to create 50,000 Captcha:&lt;/li&gt;
&lt;/ul&gt;

&lt;pre&gt;&lt;code class=&#34;language-python&#34;&gt;import random
import numpy as np
import string
from captcha.image import ImageCaptcha

image = ImageCaptcha()
clist = list(string.ascii_uppercase+string.digits)
random.shuffle(clist)

labels = []
for j in range(50000):
    randstr = &#39;&#39;
    for i in range(4):
        randstr = randstr + random.choice(clist)
    labels.append(randstr)
    data = image.generate(randstr)
    image.write(randstr, &#39;~/Desktop/Neural_nets/data_10/&#39;+randstr+&#39;.png&#39;)

with open(&#39;~/Desktop/Neural_nets/data_10/&#39;+&#39;labels.txt&#39;,&#39;w&#39;) as f:
  for labs in labels:
      f.write(&#39;%s\n&#39; %labs)
&lt;/code&gt;&lt;/pre&gt;

&lt;ul&gt;
&lt;li&gt;We are in a process of writing and training a model like: VGG16&lt;/li&gt;
&lt;/ul&gt;

&lt;div class=&#34;alert alert-note&#34;&gt;
  &lt;p&gt;Note: This post is incomplete&amp;hellip;&lt;/p&gt;

&lt;/div&gt;

</description>
    </item>
    
    <item>
      <title>Interesting Science Blogs and Websites</title>
      <link>https://arshishir.github.io/post/interesting-_website/</link>
      <pubDate>Fri, 01 Feb 2019 00:00:00 -0500</pubDate>
      
      <guid>https://arshishir.github.io/post/interesting-_website/</guid>
      <description>

&lt;p&gt;This is my attempt to keep track of interesting science related websites and blogs&lt;/p&gt;

&lt;h3 id=&#34;blogs&#34;&gt;&lt;strong&gt;Blogs&lt;/strong&gt;&lt;/h3&gt;

&lt;ol&gt;
&lt;li&gt;&lt;a href=&#34;https://johncarlosbaez.wordpress.com/&#34; target=&#34;_blank&#34;&gt;https://johncarlosbaez.wordpress.com/&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href=&#34;https://mgbukov.github.io/&#34; target=&#34;_blank&#34;&gt;https://mgbukov.github.io/&lt;/a&gt;&lt;/li&gt;
&lt;/ol&gt;

&lt;h3 id=&#34;websites&#34;&gt;&lt;strong&gt;Websites&lt;/strong&gt;&lt;/h3&gt;

&lt;ol&gt;
&lt;li&gt;&lt;strong&gt;John Carlo Baez&lt;/strong&gt;

&lt;ul&gt;
&lt;li&gt;&lt;a href=&#34;http://math.ucr.edu/home/baez/&#34; target=&#34;_blank&#34;&gt;http://math.ucr.edu/home/baez/&lt;/a&gt; &lt;br /&gt;
This webpage has many interesting topics. I have particularly gone through his exposition to information geometry (&lt;a href=&#34;http://math.ucr.edu/home/baez/information/&#34; target=&#34;_blank&#34;&gt;http://math.ucr.edu/home/baez/information/&lt;/a&gt;), which is engaging and interesting.&lt;/li&gt;
&lt;li&gt;&lt;a href=&#34;http://math.ucr.edu/home/baez/stoch_stable.pdf&#34; target=&#34;_blank&#34;&gt;http://math.ucr.edu/home/baez/stoch_stable.pdf&lt;/a&gt; &lt;br /&gt;
Quantum Techniques for Stochastic Mechanics. A book, I hope to go through one day.
&lt;br /&gt;&lt;/li&gt;
&lt;/ul&gt;&lt;/li&gt;

&lt;li&gt;&lt;p&gt;&lt;strong&gt;Gavin E. Crooks&lt;/strong&gt;&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;&lt;a href=&#34;http://threeplusone.com/gec/&#34; target=&#34;_blank&#34;&gt;http://threeplusone.com/gec/&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href=&#34;http://threeplusone.com/FieldGuide.pdf&#34; target=&#34;_blank&#34;&gt;http://threeplusone.com/FieldGuide.pdf&lt;/a&gt; (Field Guide to Continuous Probability Distributions)&lt;/li&gt;
&lt;li&gt;&lt;a href=&#34;http://threeplusone.com/on_information.pdf&#34; target=&#34;_blank&#34;&gt;http://threeplusone.com/on_information.pdf&lt;/a&gt; (On Measures of Entropy and Infromation)&lt;/li&gt;
&lt;li&gt;&lt;a href=&#34;http://threeplusone.com/Crooks-FisherInfo.pdf&#34; target=&#34;_blank&#34;&gt;http://threeplusone.com/Crooks-FisherInfo.pdf&lt;/a&gt; (Fisher Information and Statistical Mechanics)&lt;/li&gt;
&lt;li&gt;&lt;a href=&#34;http://threeplusone.com/Crooks-Whither.pdf&#34; target=&#34;_blank&#34;&gt;http://threeplusone.com/Crooks-Whither.pdf&lt;/a&gt; (Whither Time’s Arrow?)&lt;/li&gt;
&lt;/ul&gt;&lt;/li&gt;

&lt;li&gt;&lt;p&gt;&lt;strong&gt;Machine Learning&lt;/strong&gt;&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;&lt;a href=&#34;https://arxiv.org/pdf/1803.08823.pdf&#34; target=&#34;_blank&#34;&gt;https://arxiv.org/pdf/1803.08823.pdf&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href=&#34;http://www.offconvex.org/&#34; target=&#34;_blank&#34;&gt;http://www.offconvex.org/&lt;/a&gt;
&lt;br /&gt;&lt;/li&gt;
&lt;/ul&gt;&lt;/li&gt;
&lt;/ol&gt;

&lt;div class=&#34;alert alert-note&#34;&gt;
  &lt;p&gt;This is by no means a complete list. I will keeping adding.&lt;/p&gt;

&lt;/div&gt;

</description>
    </item>
    
    <item>
      <title>Journal Club</title>
      <link>https://arshishir.github.io/post/journal-club/</link>
      <pubDate>Fri, 01 Feb 2019 00:00:00 -0500</pubDate>
      
      <guid>https://arshishir.github.io/post/journal-club/</guid>
      <description>&lt;p&gt;Collections of papers that I would like to go over.&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;&lt;p&gt;&lt;strong&gt;Stochastic Thermodynamics&lt;/strong&gt;&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Arcsine laws in stochastic thermodynamics (Barato et al.)&lt;br /&gt;
&lt;a href=&#34;https://arxiv.org/pdf/1712.00795.pdf&#34; target=&#34;_blank&#34;&gt;https://arxiv.org/pdf/1712.00795.pdf&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;Thermodynamic cost of external control(Barato et al.)&lt;br /&gt;
&lt;a href=&#34;https://arxiv.org/abs/1704.03480&#34; target=&#34;_blank&#34;&gt;https://arxiv.org/abs/1704.03480&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;&lt;/li&gt;

&lt;li&gt;&lt;p&gt;&lt;strong&gt;Ecology&lt;/strong&gt;&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Constrained optimization as ecological dynamics with applications to    random quadratic programming in high dimensions (Mehta et al.)&lt;br /&gt;
&lt;a href=&#34;https://arxiv.org/pdf/1809.04221.pdf&#34; target=&#34;_blank&#34;&gt;https://arxiv.org/pdf/1809.04221.pdf&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;&lt;/li&gt;

&lt;li&gt;&lt;p&gt;&lt;strong&gt;Protein&lt;/strong&gt;&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;The strength of protein-protein interactions controls the information capacity and dynamical response of signaling networks (Wang et. al)&lt;br /&gt;
&lt;a href=&#34;https://arxiv.org/abs/1811.05371&#34; target=&#34;_blank&#34;&gt;https://arxiv.org/abs/1811.05371&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;&lt;/li&gt;

&lt;li&gt;&lt;p&gt;&lt;strong&gt;Reinforcement Learning for Physics Problems&lt;/strong&gt;&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Reinforcement Learning in Different Phases of Quantum Control (Bukov et al.)&lt;br /&gt;
&lt;a href=&#34;https://arxiv.org/pdf/1705.00565.pdf&#34; target=&#34;_blank&#34;&gt;https://arxiv.org/pdf/1705.00565.pdf&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;Reinforcement learning for autonomous preparation of Floquet-engineered states:
Inverting the quantum Kapitza  (Bukov) &lt;br /&gt;
&lt;a href=&#34;https://arxiv.org/pdf/1808.08910.pdf&#34; target=&#34;_blank&#34;&gt;https://arxiv.org/pdf/1808.08910.pdf&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;&lt;/li&gt;

&lt;li&gt;&lt;p&gt;&lt;strong&gt;Speed Limit&lt;/strong&gt;&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Geometric Speed Limit of Accessible Many-Body State Preparation (Bukov et al.)&lt;br /&gt;
&lt;a href=&#34;https://arxiv.org/pdf/1804.05399.pdf&#34; target=&#34;_blank&#34;&gt;https://arxiv.org/pdf/1804.05399.pdf&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;&lt;/li&gt;
&lt;/ol&gt;
</description>
    </item>
    
    <item>
      <title>Why?</title>
      <link>https://arshishir.github.io/post/why/</link>
      <pubDate>Fri, 01 Feb 2019 00:00:00 -0500</pubDate>
      
      <guid>https://arshishir.github.io/post/why/</guid>
      <description>&lt;p&gt;As a Ph.D. student, we are exposed to lots of new ideas and concepts. The posts on this website are my attempt to keep track of those ideas and concepts. I started a blog a couple of years ago but failed to maintain it. This is my second go at it. Better late than never.&lt;/p&gt;

&lt;p&gt;I set up this website using Hugo framework with the academic theme. I have high hopes from this set up as it natively supports $\LaTeX$.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>Unraveling the mechanism of the cadherin-catenin-actin catch bond</title>
      <link>https://arshishir.github.io/publication/plos_2018/jimm/</link>
      <pubDate>Fri, 17 Aug 2018 00:00:00 -0400</pubDate>
      
      <guid>https://arshishir.github.io/publication/plos_2018/jimm/</guid>
      <description></description>
    </item>
    
    <item>
      <title>Conformation of a flexible chain in explicit solvent: Accurate solvation potentials for Lennard-Jones chains</title>
      <link>https://arshishir.github.io/publication/jcp_2015/jimm/</link>
      <pubDate>Mon, 23 Nov 2015 00:00:00 -0500</pubDate>
      
      <guid>https://arshishir.github.io/publication/jcp_2015/jimm/</guid>
      <description></description>
    </item>
    
    <item>
      <title>Conformation of a flexible chain in explicit solvent: Exact solvation potentials for short Lennard-Jones chains</title>
      <link>https://arshishir.github.io/publication/jcp_2011/jimm/</link>
      <pubDate>Thu, 28 Jul 2011 00:00:00 -0400</pubDate>
      
      <guid>https://arshishir.github.io/publication/jcp_2011/jimm/</guid>
      <description></description>
    </item>
    
  </channel>
</rss>