<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Mukunda Selvaraj on Medium]]></title>
        <description><![CDATA[Stories by Mukunda Selvaraj on Medium]]></description>
        <link>https://medium.com/@M1029?source=rss-6a3a6c0e06b8------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*aRyXlykvL1vCmRa65cfDLA.jpeg</url>
            <title>Stories by Mukunda Selvaraj on Medium</title>
            <link>https://medium.com/@M1029?source=rss-6a3a6c0e06b8------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 04:02:54 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@M1029/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[The World Around GNU/Linux]]></title>
            <link>https://medium.com/@M1029/the-world-around-gnu-linux-fdb7c5c27aa7?source=rss-6a3a6c0e06b8------2</link>
            <guid isPermaLink="false">https://medium.com/p/fdb7c5c27aa7</guid>
            <category><![CDATA[free-software]]></category>
            <category><![CDATA[terminal]]></category>
            <category><![CDATA[gnu-linux]]></category>
            <dc:creator><![CDATA[Mukunda Selvaraj]]></dc:creator>
            <pubDate>Wed, 27 Mar 2019 23:05:42 GMT</pubDate>
            <atom:updated>2019-03-27T23:05:42.940Z</atom:updated>
            <content:encoded><![CDATA[<p>by MS1029</p><h3>What is GNU/Linux?</h3><p>It is an <strong>Operating System</strong> that manages a computer. It is entirely <strong>free</strong> and large number of developers spread over the world who always contribute for its development and existence till date. GNU/Linux is the core component and have different products which are <strong>distributions(distros)</strong> that vary with appearance and few working mechanism and functions like <strong>Internet/Networking, or Multimedia and Graphics and etc.</strong> There are hell lot of Free Software which are not just free but quality applications too. In simple term GNU/Linux is a collection of many programs, applications, libraries, developer tools, even games which talks with hardware of a computer.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/758/0*Z2Jl3oXVUW43JSDh" /></figure><p>There are many distros, Choosing the Right DISTRIBUTION of GNU/Linux, for newbies advised Distributions(Distros) are</p><pre><strong>1. </strong><a href="https://linuxmint.com/download.php"><strong>Linux Mint</strong></a></pre><pre><strong>2. </strong><a href="https://getfedora.org/en/workstation/download/"><strong>Fedora</strong></a></pre><pre><strong>3. </strong><a href="https://www.ubuntu.com/download/desktop"><strong>Ubuntu</strong></a></pre><p>The larger picture of a <strong>GNU/Linux</strong> or any <strong>Free Software </strong>are it promotes users <strong>Freedom to run, copy, distribute, study, change and improve </strong>the Application. Free software refers to four kinds of freedom,</p><pre><strong>freedom 0 </strong>:  The freedom to run the program, for any purpose.</pre><pre><strong>freedom 1 : </strong>The freedom to study how the program works, and adapt it to your needs. Access to the source code.</pre><pre><strong>freedom 2</strong> : The freedom to redistribute copies to others.</pre><pre><strong>freedom 3</strong> : The freedom to improve the program, and release your improved to public, so that the whole community benefits. And the source code should be accessed by public.</pre><p>GNU/Linux is based on <strong>Command Line Interface (CLI)</strong> and nowadays it supports both <strong>CLI and Graphical User Interface (GUI).</strong> Most of the functions like to create or delete a file/folder/directory, execute a program, install or uninstall, view documents, and etc can be done using a Terminal(also known as shell). Getting started with Terminal will ease most of your work starting from using a computer for viewing or managing database to working on a project to running or deploying a server and many more cool stuffs !!!.</p><h3>Getting started with TERMINAL</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/1*J5bg4fxaOy4atKegwMVM-A.png" /></figure><p>Large portions of GNU/Linux functionality are achieved by using the terminal. A terminal is a <strong>command-line interpreter</strong> that executes user inputted commands. <strong>Bash</strong> is a common default shell among many Linux distributions. Learning few basic commands listed below will make it easy to use any GNU/Linux OS.<br><strong>Commands (</strong>Terminal is Case-Sensitive<strong>)<br></strong>1. To Open the Terminal</p><pre>$ Ctrl + Alt + T<br>$ Ctrl + Shift+ T  (To open terminal in tab fashion)</pre><p>2. To end a execution</p><pre>$ Ctrl + C</pre><p>3. Clearing the Terminal Screen</p><pre>$ clear <br>or <br>$ Ctrl + L</pre><p>4. To Auto-complete files or directory names</p><pre>TAB (press Tab button)</pre><p>5. To show all possible files or directory</p><pre>TAB TAB (press Tab button twice</pre><p>6. For any Help or Information on Commands</p><pre>$ man &lt;command name&gt; (To quit a man page use &#39;q&#39; button)<br>or<br>$ &lt;command name&gt; --help</pre><p>7. ls stands for <em>list</em>.</p><pre>$ ls<br>$ ls -a  (lists all the files)<br>$ ls -l  (lists details like owner, date, type, etc.)<br>$ ls - *.pdf  (lists all the files of type pdf, not just pdf it can be other type extensions too)</pre><p>8. cd is Change Directory which allows to traverse through directories and file paths</p><pre>$ cd <br>or <br>$ cd ~  (Changes the current directory to Root Directory)<br>$ cd -  (Changes the current directory to previous directory)<br>$ cd &lt;dir&gt;  (Changes the current directory to directory &lt;dir&gt;)</pre><p>9. To display the Current directory</p><pre>$ pwd  (Working Directory)</pre><p>10. To Copy file/Directories</p><pre>$ cp &lt;Source&gt; &lt;Destination&gt;  (Copies file from source to destination)<br>$ cp file1 file2 file3 ...  dir  (Copies the following files to &lt;dir&gt;)<br>$ cp -i  (Ask user confirmation if the target file already exist)<br>$ cp -p  (preserves the attributes such as time, mode, ownership of the source file)</pre><p>11. To Move file</p><pre>$ mv &lt;old_name&gt; &lt;new_name&gt;  (Renames a given file or directory)<br>$ mv -i  (Ask user confirmation if the new file already exists)<br>$ mv -p  (preserves the attributes such as time, mode, ownership of the source file)<br>$ mv -f  (Force the move by overwriting the Destination file)</pre><p>12. To Remove a file(Delete)</p><pre>$ rm file1 file2 file3 file4  (Removes the given files)<br>$ rm -i  (Ask for user confirmation)<br>$ rm -r dir1 dir2 dir3  (Removes the given Directories including contents)<br>$ rm -f  (Force remove)</pre><p>13. Create or Remove Directories</p><pre>$ mkdir dir1 dir2 dir3  (Creates directories with given names)<br>$ touch &lt;file_name&gt;</pre><pre>$ rmdir dir1 dir2 dir3  (Removes the given directories only when they are empty)<br>$ rm -r &lt;dir&gt;  (Removes the given directories and no need to be empty)<br>$ rm -rf &lt;dir&gt;  (Removes the given directories without asking)</pre><p>14. Displaying Contents of files</p><pre>$ cat file1 file2 file3  (Concatenates and outputs the contents of the files)<br>$ vi &lt;file_name&gt;  (Edit the given file)<br>$ wi  (Gives the count of line-words-character) <br>$ wi -l  (Gives the count of line/s in the file) <br>$ wi -w  (Gives the count of word in the file) <br>$ wi -c  (Gives the count of character in the file)</pre><p>15. To Search files</p><pre>$ grep &lt;pattern&gt; &lt;file&gt;</pre><p>16. Ordering</p><pre>$ sort &lt;file&gt;   (Sorts the lines in the file in Character order and outputs them)<br>$ sort -r &lt;file&gt;  (Sorts the lines in reverse order in the file in Character order and outputs them)</pre><p>17. Echo</p><pre>$ echo &lt;option&gt; &lt;String&gt;  (Strings should be represented using double quotes)</pre><p>18. Find</p><pre>$ find &lt;dir&gt; &lt;file_name&gt;</pre><p>19. To display the current User</p><pre>$ whoami</pre><p>20. Similarly date, calendar, time, and others can be used</p><pre>$ date<br>$ calendar<br>$ time</pre><p>21. Compressing and Uncompressing</p><pre>$ gzip &lt;options&gt;&lt;file&gt;<br>$ gunzip &lt;options&gt;&lt;file&gt;</pre><p>22. To close terminal</p><pre>Ctrl + Shift + W  (terminal tab)<br>Ctrl + Shift + Q  (entire terminal)</pre><h3>Conclusion</h3><p>When we have Free Software for very prominent proprietor software which gives the same or better experience, <strong>Why not use it ?</strong> The word free in Free Software is <strong>free as in Freedom</strong>, the source code is open to all. Few may feel the source code is open, any one can place a malware and upload, it is reasonable. But the truth is for every free software there is a<strong> Community of developers</strong> working on it and developers are assigned roles and any new updates or upgrades, needs to be reviewed by them(prominent person working in the community like editors, reviewer, etc). So whenever someone tries to upload a malware or such, its removed when developers detect them.</p><p>So people working on GNU/Linux get the best use and experience Free Software. Most importantly people interested in <strong>Free Software and its philosophy</strong> need not contribute only by coding and development but <strong>spreading the knowledge</strong> of Free Software to our surrounding is also a big contribution for a better <strong>FREE DIGITAL SOCIETY</strong> that we live in.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=fdb7c5c27aa7" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[TensorFlow basics]]></title>
            <link>https://medium.com/@M1029/tensorflow-3703e41d2d97?source=rss-6a3a6c0e06b8------2</link>
            <guid isPermaLink="false">https://medium.com/p/3703e41d2d97</guid>
            <category><![CDATA[linear-regression]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[tensorflow]]></category>
            <category><![CDATA[basics]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <dc:creator><![CDATA[Mukunda Selvaraj]]></dc:creator>
            <pubDate>Sun, 15 Apr 2018 08:26:12 GMT</pubDate>
            <atom:updated>2018-07-09T08:06:53.292Z</atom:updated>
            <content:encoded><![CDATA[<p>TensorFlow is an open source free software and frame work for machine learning and deep learning and flexible numerical computing. TensorFlow is mainly designed for deep neural network models. It was developed by researchers and engineers from the Google Brain team within Google’s AI organization.</p><p>A <strong>tensor</strong> is a generalization of scalars(0D), vectors(1D), matrices(2D), 3d tensor(3D) and so on till nD tensor(nD) of potentially higher dimensions. Internally, TensorFlow represents tensors as n-dimensional arrays of base datatypes.</p><p>A tf.Tensor has the following properties:</p><ul><li>a data type (float32, int32, or string)</li><li>a shape</li></ul><p>When writing a TensorFlow program, the main object you manipulate and pass around is the tf.Tensor.</p><p>Some types of tensors are</p><ul><li>tf.constant</li><li>tf.placeholder</li><li>tf.SparseTensor</li><li>tf.Variable</li></ul><p>Using eager execution, we can excel in TensorFlow in very less duration. On restrictions of computational limitations and to overcome them on bandwidth memory locations limits and others, we need to use hadoop(elephant).</p><h4><strong>Data representation</strong></h4><p>The various data representations that can be used in TensorFlow are</p><ul><li><a href="https://www.tensorflow.org/tutorials/wide">TensorFlow Linear Model</a> uses columns to feed a variety of data types to linear model in order to solve a classification problem.</li><li><a href="https://www.tensorflow.org/tutorials/wide_and_deep">TensorFlow Wide &amp; Deep </a>Learning, which builds on the TensorFlow linear model adding a deep feed-forward neural network component and a DNN-compatible data representation.</li><li><a href="https://www.tensorflow.org/tutorials/word2vec">Vector Representations of Words</a>, which creates an embedding for words.</li><li><a href="https://www.tensorflow.org/tutorials/kernel_methods">Improving Linear Models Using Explicit Kernel Methods</a>, which improves the quality of a linear model by using explicit kernel mappings method.</li></ul><h4>So here are few basics to be followed:</h4><ol><li><strong>In order to import tensorflow.</strong></li></ol><pre>import tensorflow as tf</pre><p><strong>2. In order to define a tensorflow constant.</strong></p><pre>a = tf.constant(1)</pre><p><strong>3. In order to print a tensorflow.</strong></p><pre>print(type(a))<em>[here the datatype is tensor]</em></pre><p><strong>4. Creating arrays and converting them to tensors</strong>.</p><pre>b = tf.constant([1, 4, 787, 998])</pre><p><strong>5. In order to print the tensor array type.</strong></p><pre>print(type(b))</pre><p>[T<em>ensorflow does not evaluate the tensor until you create a session and run them</em>.]</p><p><strong>6. Evaluation of tensors hence create a session and run them.</strong></p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/5c650f06733fbf14d80c66d49fda99f7/href">https://medium.com/media/5c650f06733fbf14d80c66d49fda99f7/href</a></iframe><p>7. <strong>Computation Graphs<br></strong>Computations are structured in the form of graph in Tensorflow.<br>lets consider <strong>a, b </strong>and<strong> c</strong> are variables.</p><p><strong>J = 3 ( a + b * c)</strong></p><blockquote><strong>b * c = u <br>a+ u = v</strong><br><strong>J = 3 * v</strong></blockquote><p>It is a graph representation due to</p><ul><li>parallelism</li><li>distributed execution</li><li>compilation</li></ul><h4>Few examsples:</h4><ol><li><strong>Print hello world</strong></li></ol><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/42f057251146bdf4fbdbc6ddda6da309/href">https://medium.com/media/42f057251146bdf4fbdbc6ddda6da309/href</a></iframe><p><strong>Output</strong> : Hello Wolrd !</p><p>2. <strong>Tensorflow placeholders</strong></p><p>dictionary datatype { var: ‘statements’}</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/ffb27a2c2e810c9b109b675a1497bff3/href">https://medium.com/media/ffb27a2c2e810c9b109b675a1497bff3/href</a></iframe><p><strong>Output<em>:</em></strong> Hello People</p><p>3<strong>. For multiple placeholders</strong></p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/0425d0dfc47750a4f52b54fc32916521/href">https://medium.com/media/0425d0dfc47750a4f52b54fc32916521/href</a></iframe><p><strong>Output :</strong> [array(‘People’, dtype=object), array(456, dtype=int32), array(10.2, dtype=float32)]</p><p>4. <strong>Create two placeholders and multiply them by feeding values for them and run</strong></p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/782e037eec70a4549323131e248a78ad/href">https://medium.com/media/782e037eec70a4549323131e248a78ad/href</a></iframe><p><strong>Output </strong>: 32</p><h3><strong>Operations</strong></h3><p>There are several operations dealt with TensorFlow and please find them in these below link.<br><a href="https://www.tensorflow.org/api_guides/python/math_ops">Operations on Tensorflow</a><br>[credits:TensorFlow Website]</p><h3>Linear Regression using TensorFlow</h3><p>Here we will be dealing with Linear models, in order to do so we need to train the model with Linear regression or logistic regression. Let’s start with following and build a TensorFlow model to predict the price of land using size and price from a given census data.</p><ol><li>Reading the data</li><li>Converting data into tensors</li><li>Selection of the model</li><li>Defining the Linear regression model</li><li>Train the model</li><li>Evaluate the model</li><li>Normalisation to overcome Overfitting</li></ol><ul><li><strong>import</strong></li></ul><p>sklearn is a library used for machine leaning library</p><ul><li><strong>split</strong></li></ul><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/06345a460b4bba3894853f86667c94c4/href">https://medium.com/media/06345a460b4bba3894853f86667c94c4/href</a></iframe><ul><li><strong>normalise</strong></li></ul><p>array.std()- standard deviation of array</p><p>(array — array.mean()) / array.std() this gives a cosian fucntion curve</p><p>so wht we do is (array — array.mean)/array.std = data normalisation</p><ul><li><strong>scatter plot</strong></li></ul><p>%matplotlib inline<br>import matplotlib.pyplot as plt</p><p>again normalise</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/9b98de3cb965a71f01bb00c723b820e5/href">https://medium.com/media/9b98de3cb965a71f01bb00c723b820e5/href</a></iframe><ul><li><strong>model function</strong></li></ul><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/c13544eca6bb05976c71c0ff67ef87b0/href">https://medium.com/media/c13544eca6bb05976c71c0ff67ef87b0/href</a></iframe><p>model gives prediction</p><ul><li>we are taking 2n for the sake of simpler mathematics</li><li>training_iteration = 200<br>done we are gredient design is run many times to get to the optimal</li><li>zip is for loop which runs</li></ul><p><strong>Train and evaluation of model</strong></p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/bb86a0ca192b92d635c1968fa4c222d3/href">https://medium.com/media/bb86a0ca192b92d635c1968fa4c222d3/href</a></iframe><p>Model has been normalised and the below plot indicates the difference between after and before normalisation.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/761/1*lupFHpakvFSLmk1urOJfBg.png" /><figcaption>After Normalisation and before Normalisation</figcaption></figure><p>I have pasted the entire model here:</p><p><a href="https://colab.research.google.com/drive/1ZEF-fSqjWdDgvpNazddr01a7qQGlRtGu#scrollTo=ocSoX5Zy9kDg">Linear_Model_Regression</a></p><h3><strong>Conclusion : Why TensorFlow?</strong></h3><p>It is a backend framework, where we use it for complex problems. It do is easier to make complex problems to easier ones through tensorflow by virtue of using linear regression , logical regression and so on.<br>Prediction done by TensorFlow are used in large scale such as early stage cancer deduction, Image recognition and Video Detection etc.</p><p>AI required huge amount of deep learning algorithm, computing and lots parameters. Whereas implementation of TensorFlow made it hassle-free. So TensorFlow is powerful and well documented.</p><p>It is also changing the way developers are interacting with machine learning technology. It brings down the errors, and training the model is easier too.</p><p>To Predict, Train It, Inorder to Gain Momentum Through <strong>TENSORFLOW</strong></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3703e41d2d97" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>