<?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 Yongrae Kyle Jo on Medium]]></title>
        <description><![CDATA[Stories by Yongrae Kyle Jo on Medium]]></description>
        <link>https://medium.com/@dreamgonfly?source=rss-28d9b84cbbeb------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*sIQVMP0go1QVjwNq.jpg</url>
            <title>Stories by Yongrae Kyle Jo on Medium</title>
            <link>https://medium.com/@dreamgonfly?source=rss-28d9b84cbbeb------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 15:09:18 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@dreamgonfly/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[Introduction to Machine Learning and Python]]></title>
            <link>https://medium.com/@dreamgonfly/introduction-to-machine-learning-and-python-eea555d0fde8?source=rss-28d9b84cbbeb------2</link>
            <guid isPermaLink="false">https://medium.com/p/eea555d0fde8</guid>
            <dc:creator><![CDATA[Yongrae Kyle Jo]]></dc:creator>
            <pubDate>Sat, 12 Jul 2014 11:38:59 GMT</pubDate>
            <atom:updated>2015-12-29T09:22:52.668Z</atom:updated>
            <content:encoded><![CDATA[<h4>Machine Learning with Python 0</h4><blockquote>“Well begun is half done”</blockquote><p>This post is for anyone who is curious about machine learning but is new to Python programming language or programming in general. After reading this you will get some ideas about what machine learning and Python mean, and why Python is a good way to approach machine learning especially for beginners.</p><h3>Why Machine Learning</h3><p>Machine Learning is the idea that a computer can decide and do something based on data without a person directly instructing what to do for it. As experience is the data necessary for a human being to learn, data is necessary for a computer to learn as well. Just as the brain analyzes patterns in experiential data and makes appropriate decisions based off it, a computer can find interesting patterns in sets of data and adjust its programming accordingly.</p><p>Let’s take the ability to recognize hand-written digits for example. For a human being, it is a piece of cake. However, as all handwriting differs from each other, it is a hard task for a computer to tell which figure means which number. To make matters worse, for human beings themselves it’s difficult to explain the process by which digits are recognized, because it just happens unconsciously inside the brain. As a result, it’s very difficult to teach a computer the proper instructions on how to identify handwritten digits.</p><p>However, if enough data on hand-written digits and their corresponding number is collected and analyzed, a computer can learn the process behind recognizing hand-written digits. In school, studying questions and answers repetitively helps students solve similar questions well. It is very similar to how machine learning works.</p><p>One distinctive competence of machine learning is that an algorithm for one problem can be applied to other problems with different set of data. For example, the handwriting recognition algorithm also can be used for auto-driving, human face detection, and voice recognition.</p><p>Today, machine learning can be seen everywhere. Search engines like Google and Bing, the photo tagging application of Facebook and Apple, and Gmail’s spam filtering are all everyday examples of machine learning.</p><h3>Why Python</h3><p>Python is simple yet powerful programming language. Python’s syntax is succinct and human-friendly, so it is famous for being easy to read and write. This is why Python is one of the most recommended programming language for beginners. One seeing is more worth than thousand hearings. Take a look at a line of Python code below.</p><pre>if 4 in [1,2,3,4]: print (“There is 4&quot;)</pre><p>Above code can be read like this : ‘If 4 is in the list of 1, 2, 3, 4, then print “There is 4”’. Even if you don’t know anything about Python, you can read the above line and grasp the meaning of the sentence instinctively. It’s like reading an English sentence.</p><p>Programming is a conversation between a human being and a computer in a mutually understandable language. If you can talk to a computer in a language familiar to you, you would not have to worry about the way a computer thinks so you could rapidly express your thoughts to a computer and could easily read others’ thoughts. That’s exactly what Python is all about.</p><p>Python is simple, but that doesn’t mean it sacrifices its power. Python comes with excellent functionalities such as graph drawing, scientific research, web connectivity, game development, and so on. Python is actively used in industry, science, and internet around the world. Dropbox is made in Python. About half of Google is made in Python. Even games like Sid Meier’s Civilization IV. You can do almost anything with Python that you could do in any other programming language.</p><p>If you do know some of the other programming languages, you might well understand these terms describing Python: interpreted, object-oriented, and dynamic. As an interpreted language, Python helps you interact with it line by line while programming. As an object-oriented language, Python treats everything as objects so they are easily reusable. As a dynamic language, Python determines the type of variables on behalf of you and allows objects to be changed on the fly. Each of these Python’s properties contributes to facilitating rapid development.</p><p>Python was created by Guido Van Rossum in the late 1980s. It is a successor to ABC programming language, which was also interactive, dynamic, and highly focused on raising programming literacy among people. ABC itself was a successor to SET Language, which was based on highly mathematical concepts. Python succeeded philosophies of these languages, so it aims for both easy and high-functional language.</p><p>Just for information, there are two major versions of Python: Python2 and Python3. Python3 dropped some redundant features of Python2 with less regard for backwards compatibility. There is controversy over which version of Python you should use. Python3 is cleaner and more consistent, but Python lectures tended to recommend Python2 because many useful modules made in Python2 could not be used in Python3. However, as time passed, most of Python2 modules have been ported to Python3, so there is no difficulty to use Python3 anymore, especially for beginners who do not need sophisticated modules. Therefore, we’re going to be discussing Python3 in this series of lectures.</p><h3>Why Machine Learning with Python</h3><p>Learning a programming language is like learning how to ride a bicycle. You can’t really learn it by only reading about it or studying theories. When I first started learning Python, I myself had spent many months only reading up on specification of Python syntax, which I have almost forgotten now. It was only when I strove to make my own program that I actually learned how to code Python by heart. If you want to learn Python, you need to confront real problems. Taking on the challenge of machine learning problems can be one way to facilitate learning Python.</p><p>The same is true for machine learning itself. By trying to implement a machine learning algorithm on your computer, you can understand the logic behind the scenes more deeply. Also, your own knowledge and codes will be able to be applied to other problems.</p><p>Of course, Python is not the only language that can be used for machine learning. There are plenty of programming languages that can implement machine learning algorithms. Two most competent options are Matlab and R.</p><p>Matlab is a high-level programming language heavily focused on handling matrix. It is so popular in numerical computing that Python was first intended to imitate Matlab in handling matrix. R is getting popularity in statistics and data analysis. So why does this lecture discuss Python? The most important reason is that Python is easy, as mentioned above, but there are two more reasons.</p><p>First, Python is free. Matlab is expensive. It costs more than most individuals can afford. Although Matlab provides high-performing functionalities, they are not needed unless you are a serious researcher in academia or industry. For beginners, Python is a pretty good free alternative to Matlab.</p><p>Second, Python is a general-purpose language. Matlab and R are only for computing and data analysis. That means, you can easily combine data analysis with other functions in Python. The other functions include collecting data from the web, making a program use the result, or directly reflecting the result on a website. Moreover, you can use Python skills learned while solving machine learning problems for your own purposes other than data analysis.</p><h3>References</h3><p>If you are interested and want further reading, check out the links below. Some of them are where I myself studied Python and machine learning. Unfortunately, many lectures on Python are still about Python2. Again, time has passed, and now Python3 is the present and future of the language. I strongly recommend Python3 for beginners unless you have obvious reasons to use Python2. But, this list includes some good lectures in Python2, because learning Python2 doesn’t mean you can’t learn Python3. The list is ordered by accessibility.</p><ul><li><a href="http://opentutorials.org/course/84/587">Coding Everybody’s video lectures on Python</a> (Korean, Python2)</li></ul><p>If you know literally nothing about Python, you should first watch these short videos. Each video is less than 15mins. These lectures don’t go deep into Python, but it’s enough to get some basic concepts of it.</p><ul><li><a href="http://www.codecademy.com/">Codecademy</a> (Python2)</li></ul><p>In this revolutionary platform, you can learn Python by interactively writing codes by instructions.</p><ul><li><a href="https://wikidocs.net/book/1">Jump to Python</a> (Korean)</li></ul><p>This is Well-organized and well-explained lectures about Python. Its lecture style is somewhat typical, but it doesn’t miss details of Python.</p><ul><li><a href="http://learnpythonthehardway.org/book/">Learn Python The Hard Way</a> (Python 2)</li></ul><p>This lectures focus on engraving details of Python in your mind. It’s a hard way, but you can better remember what you learn in the hard way.</p><ul><li><a href="http://www.diveintopython3.net/">Dive into Python</a></li></ul><p>I really like this insightful style of teaching Python. But this lectures jump into advanced concepts of Python shortly after skimming basic grammar. I would recommend this lecture for people who are already accustomed to basic concepts of Python.</p><ul><li><a href="http://victoria.lviv.ua/html/fl5/NaturalLanguageProcessingWithPython.pdf">Natural language Processing with Python</a> (Python2)</li></ul><p>This is an excellent way to learn both natural language processing and Python at the same time.</p><ul><li><a href="http://www.manning.com/pharrington/">Machine Learning in Action</a></li></ul><p>This book is about implementing machine learning with Python. However, it doesn’t explain much about Python. So you have to be experienced in Python before reading this book.</p><ul><li><a href="https://www.coursera.org/course/ml">Andrew Ng’s Machine Learning course on Coursera</a> (not Python)</li></ul><p>The single best lecture on machine learning I’ve ever known. This lecture uses Matlab in assignments, but there is Python version of the assignments created by students.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=eea555d0fde8" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Economic growth no longer creates job opportunities]]></title>
            <link>https://medium.com/@dreamgonfly/economic-growth-no-longer-creates-job-opportunities-30d0c14a98a0?source=rss-28d9b84cbbeb------2</link>
            <guid isPermaLink="false">https://medium.com/p/30d0c14a98a0</guid>
            <dc:creator><![CDATA[Yongrae Kyle Jo]]></dc:creator>
            <pubDate>Tue, 25 Feb 2014 23:08:09 GMT</pubDate>
            <atom:updated>2014-02-25T23:08:09.085Z</atom:updated>
            <content:encoded><![CDATA[<p>The Luddite movement failed in 19th century. Their arguments that technology destroy job places were fooled as they are called ‘Luddite fallacy’. There are compelling evidences that show advance in technology results in more robust economy, creating more job opportunities. So many economists firmly believe technology is one of the main factor of job creation.</p><blockquote>“Indirect effects occur when technological improvements increase demand and lower the prices consumers pay their for goods and services, thereby giving them more purchasing power and stimulating growth in other sectors, which leads to a self-reinforcing economic expansion.” <em>— </em><a href="http://www2.itif.org/2013-are-robots-taking-jobs.pdf">Are Robots Taking Our Jobs, or Making Them?</a></blockquote><p>It seemed that this claim holds true for a while. However, then comes a picture.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/856/1*tuKBJSQ1DDDs-feH4a2daA.png" /><figcaption>U.S. productivity and employment (index: 1947 = 100)</figcaption></figure><p>A picture is worth a thousand words. In above picture, it is clear that since 2000 employment has no longer increased while productivity has been at a record high. It shows that technological development now doesn’t mean expansion of job market anymore. Why this happens?</p><p>Some might say that this is because technology replaced people. Of course, it’s obvious that some technology takes the places where people used to work. However, technology also create new jobs. As a matter of fact, technology has created more jobs than it destroyed until 20 years ago, as the above picture shows. So, just saying that technology destroyed jobs cannot be an answer. Right question might be ‘What’s the differences between the past and today?’</p><p>Paul Krugman, a famous economist and columnist, answered this question with two terms — <a href="http://www.nytimes.com/2012/12/10/opinion/krugman-robots-and-robber-barons.html?_r=0">Robots and Robber Barons</a>. Robots, automation, and software are substituting not only low-skilled workers, but also sophisticated, intellectual workers. For an example, Google is experimenting auto-driving cars, which can drive better than humans. This is very different trend from the age of Industrial Revolution, where technology displaced only physical and repetitive workers. As technology develops further, this phenomenon would become more serious.</p><p>Second, Monopoly power can be an huge factor of today’s unemployment, as Krugman said:</p><blockquote>What about robber barons? We don’t talk much about monopoly power these days; antitrust enforcement largely collapsed during the Reagan years and has never really recovered. Yet <a href="http://www.washingtonmonthly.com/features/2010/1003.lynn-longman.html">Barry Lynn and Phillip Longman</a> of the New America Foundation argue, persuasively in my view, that increasing business concentration could be an important factor in stagnating demand for labor, as corporations use their growing monopoly power to raise prices without passing the gains on to their employees.</blockquote><p>To sum up, Technology has been spurring productivity, replacing job opportunities for people. Corporations have earned a lot of profits by introducing automation, at a labor’s expense. Now, technology no longer creates job opportunities and economic expanse doesn’t guarantee high employment anymore. We are in the age of ‘jobless growth’.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=30d0c14a98a0" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>