test.html

<h1>Learning Python for Machine Learning</h1>
<p>Machine learning is a branch of Artificial Intelligence, or AI, that focuses on building applications that learn from data and improve their accuracy with experience. Algorithms are built and trained and identify patterns in massive amounts of data, that could be used to make decisions and predictions for new sets of data. Machine learning has become an integral part of technology today. From robot vacuums, to digital assistants such as Amazon&rsquo;s Alexa and Apple&rsquo;s Siri, to medical imaging analysis systems, machine learning has a wide range of applications. With big data becoming increasingly popular and larger, computing become more affordable and powerful, machine learning has potential of driving greater efficiency in our work and personal lives.</p>

<h2>Why Python?</h2>

<p>For machine learning, Python is the most widely used programming language, and for good reason! Python is a high-level, dynamic, object-oriented programming language. It is claimed to be the fastest growing language today. The open-source language is easy to learn for beginners and experts alike. It is known for offering higher productivity with lesser code, compared to other languages. Python thrives with the various libraries created to enhance the Python language. These libraries have allowed Python to perform several functionalities seamlessly. Alongside its libraries, Python offers a versatile range of applications, such as data science and visualization, web development, game development, desktop GUI, web-scraping, machine learning and artificial intelligence among many others. Python is most definitely a very sought-after skill across several industries and occupations.</p>

<p>There are several great resources available online - free and paid - that offer world-class teaching of Python. The plethora of resources allows you to find a teaching method suitable to you. While some may learn better with simple Youtube videos, some may prefer a textbook-style learning method. Identifying the method that works best for you will help you be successful in learning Python efficiently.</p>

<h2>Download Python</h2>

<p>If you have not already, the first step is to download Python on to your computer at:</p>

<p>https://www.python.org/downloads/</p>

<p>Python is a cross-platform language, meaning it does not differentiate between Linux, or Windows or Macintosh systems. Files made on a Windows system could be run on Linux and Macintosh systems, and vice versa, with no compiling step required.</p>

<h2>Where to code?</h2>

<p>There are several platforms that could be used to develop python code. It is important to find a platform that is most comfortable for you, and offers the functionalities you need.</p>

<ul>
	<li>Google Colab is a web-based notebook environment that allows one to write and execute code in exclusively Python. No download is required, and you can code straight from your browser. One restriction of Colab is that Internet connection is required to work.&nbsp; One advantage of Google Colab is that it is tied to various Google Services: for example, you can store Colab notebooks on your Google Drive for easy access.</li>
	<li>Jupyter is an open-source web-based notebook environment for several programming languages, not limited to Python. A quick installation is required. The advantage of Jupyter is that it functionality for several programming languages, making it a one-stop source for all one&rsquo;s coding needs.</li>
</ul>

<p>Google Colab and Jupyter are very similar in functionality. They both have the ability to import and export to python files to GitHub, and other platforms.</p>

<p>&nbsp;</p>

<ul>
	<li>IDE &ndash; There are several downloadable interactive development editors that can be used for Python.&nbsp; These are generally applications downloaded to your computer. Some IDE&rsquo;s for python include:</li>
</ul>

<p>&bull;&nbsp; PyCharm</p>

<p>&bull;&nbsp; Atom</p>

<p>&bull;&nbsp; Spyder</p>

<p>&bull;&nbsp; Visual Studio Code</p>

<p>&bull;&nbsp; Kite</p>

<p>&bull;&nbsp; and many more.</p>

<h2>How to learn Python basics</h2>

<p>After downloading Python, the next step is to start learning the basics of the Python language. Having a grasp of the syntax, various data types, data structures and functions of python would make the coding process seamless and intuitive. The python website offers detailed documentation that can be easily navigated through to carry out various functions.</p>

<p>https://docs.python.org/3/&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;</p>

<p>MIT offers their &lsquo;Introduction to Computer Science and Programming in Python&rsquo; course lectures on YouTube for free.&nbsp; The course is geared to students with little or no programming experience.&nbsp; Their website offers a set of programming assignments that are great practice in your learning process.</p>

<p>Youtube Playlist: <a href="https://www.youtube.com/playlist?list=PLUl4u3cNGP63WbdFxL8giv4yhgdMGaZNA">https://www.youtube.com/playlist?list=PLUl4u3cNGP63WbdFxL8giv4yhgdMGaZNA</a></p>

<p>Course Website:</p>

<p><a href="https://ocw.mit.edu/courses/electrical-engineering-and-computer-science/6-0001-introduction-to-computer-science-and-programming-in-python-fall-2016/index.htm">https://ocw.mit.edu/courses/electrical-engineering-and-computer-science/6-0001-introduction-to-computer-science-and-programming-in-python-fall-2016/index.htm</a></p>

<p>University of Michigan offers a Coursera course, &lsquo;Python for Everybody&rsquo; to introduce the basics of Python. The content is available for free. However, one can gain access to graded assignments and receive certification as a paid feature. An advantage of this Coursera course is that it offers a learning structure and weekly deadlines.</p>

<p>Coursera Course:</p>

<p><a href="https://www.coursera.org/learn/python">https://www.coursera.org/learn/python</a></p>

<p>For those who prefer a textbook-reading style resource, &lsquo;Python for Absolute Beginners&rsquo; is ideal. This open-source guidebook also acts as a great reference to refer back to if you ever need extra help when coding.</p>

<p>Textbook:</p>

<p><a href="http://index-of.es/Python/Python%203%20for%20Absolute%20Beginners.pdf">http://index-of.es/Python/Python%203%20for%20Absolute%20Beginners.pdf</a></p>

<p>Github Resource for Textbook:</p>

<p><a href="https://github.com/Apress/python-3-for-absolute-begs">https://github.com/Apress/python-3-for-absolute-begs</a></p>

<h2>Python Libraries</h2>

<p>Along with the basics of Python, it is important to learn the most widely used python libraries. Python libraries help supplement the coding experience by providing increased functionality. Some of the most popular libraries include:</p>

<ul>
	<li>Numpy is a Python library to support large, multi-dimensional arrays and matrices, along with a large collection of high-level mathematical functions to operate on these arrays.</li>
	<li>Pandas is a Python library used for data manipulation and analysis. In particular, it offers data structures and operations for manipulating numerical tables and time series.</li>
	<li>Matlibplot is a plotting library for the Python programming language and its numerical mathematics extension NumPy.</li>
</ul>

<p>The following handbook for Python Data Science is a great resource for learning the various functionalities for the three main libraries:</p>

<p>Handbook:</p>

<p><a href="https://colab.research.google.com/github/jakevdp/PythonDataScienceHandbook/blob/master/notebooks/Index.ipynb">https://colab.research.google.com/github/jakevdp/PythonDataScienceHandbook/blob/master/notebooks/Index.ipynb</a></p>

<h2>Machine Learning Algorithms</h2>

<p>Once you have learned the basics of Python, you can start with learning machine learning basics and tactics. Knowing the basic algorithms and key concepts of machine learning acts as a foundation for creating ML programs. There are several commonly used algorithms that are important to know. The machine learning process generally makes up of data processing, choosing an algorithm, model tuning, training and testing. Knowing the dataset is very crucial for choosing an algorithm. Having an idea of what is required of the machine learning application guides the choice of which algorithm is used.</p>

<p>For starters, Chapter 5 of the Python Data Science Handbook offers a great snapshot of the most common machine learning algorithms and their real-life applications.</p>

<p>Handbook:</p>

<p><a href="https://colab.research.google.com/github/jakevdp/PythonDataScienceHandbook/blob/master/notebooks/05.00-Machine-Learning.ipynb#scrollTo=OTaNmOYY9hJA">https://colab.research.google.com/github/jakevdp/PythonDataScienceHandbook/blob/master/notebooks/05.00-Machine-Learning.ipynb#scrollTo=OTaNmOYY9hJA</a></p>

<p>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; For a more in-depth teaching of machine learning algorithms, Andrew Ng&rsquo;s Coursera course, &lsquo;Machine Learning&rsquo; from Stanford. Andrew Ng is considered a pioneer in artificial intelligence, machine learning and deep learning. The course videos are also present on YouTube for convenient, easy access.&nbsp;</p>

<p>Coursera Course:</p>

<p><a href="https://www.coursera.org/learn/machine-learning#syllabus">https://www.coursera.org/learn/machine-learning#syllabus</a></p>

<p>Youtube Playlist:</p>

<p><a href="https://www.youtube.com/playlist?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN">https://www.youtube.com/playlist?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN</a></p>

<h2>Machine Learning Libraries</h2>

<p>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; There are several machine learning libraries that simplify the process of programming for machine learning. It is now possible to build the entire machine learning process of data processing, tuning, training and testing with very few lines of code. The availability of several such libraries is what makes Python very desirable in the field of machine learning. In addition to Numpy, Pandas, Matlibplot, some of the common machine learning libraries include:</p>

<ul>
	<li><strong>Scikit-Learn</strong>, one of the most common ML libraries, is most apt for the classical machine learning algorithms. It is most ideal for the novice programmers.</li>
</ul>

<p><a href="https://scikit-learn.org/stable/">https://scikit-learn.org/stable/</a></p>

<p><a href="https://www.journaldev.com/18341/python-scikit-learn-tutorial">https://www.journaldev.com/18341/python-scikit-learn-tutorial</a>&nbsp; &nbsp; &nbsp; &nbsp; &nbsp;&nbsp;</p>

<ul>
	<li><strong>TensorFlow</strong> is the most common library for deep learning. It facilitates the creation of a multi-layered nodal system, often used for setting up, training, and deploying artificial neural networks with large datasets.</li>
</ul>

<p><a href="https://www.tensorflow.org">https://www.tensorflow.org</a></p>

<p><a href="https://www.journaldev.com/18175/python-tensorflow-tutorial">https://www.journaldev.com/18175/python-tensorflow-tutorial</a></p>

<ul>
	<li><strong>PyTorch</strong> is a python-based scientific computing package targeted for Machine Learning. It is a replacement for NumPy and provides maximum speed and flexibility by making use of multiple GPUs.</li>
</ul>

<p><a href="https://pytorch.org">https://pytorch.org</a></p>

<p><a href="https://www.journaldev.com/36558/getting-started-with-pytorch">https://www.journaldev.com/36558/getting-started-with-pytorch</a></p>

<ul>
	<li><strong>Keras</strong> is a very popular Machine Learning library for Python. It is a high-level neural networks API capable of running on top of TensorFlow.</li>
</ul>

<p><a href="https://keras.io">https://keras.io</a>&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;</p>

<p><a href="https://www.journaldev.com/18314/keras-deep-learning-tutorial">https://www.journaldev.com/18314/keras-deep-learning-tutorial</a></p>

<p>Once you have grasped the python and machine learning concepts, let&rsquo;s walk through an example of a machine learning project in python! See how we can put all of our learned concepts together to generate a handwriting character recognition program. LINK TO BLOG POST</p>

<p>When you are ready to develop your own machine learning projects, here are some tools to help you get started! You can also check out our blog on executing a machine learning project here: LINK TO BLOG POST</p>

<h2>What is Kaggle?</h2>

<p><strong>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; </strong>After you have learned the basics of Python and machine learning, it is time to start implementing your own projects. Kaggle allows users to find and publish data sets, explore and build models in a web-based data-science environment, work with other data scientists and machine learning engineers, and enter competitions to solve data science challenges. With the vast availability of datasets on Kaggle from various fields, such as health, business, economics, entertainment, and education, it is easy to tailor your machine learning project to your interests.</p>

<p><a href="https://www.kaggle.com">https://www.kaggle.com</a></p>

<h2>What is GitHub?</h2>

<p>At a high level, GitHub is a website and cloud-based service that helps developers store and manage their code, as well as track and control changes to their code.</p>

<p><a href="https://github.com">https://github.com</a></p>

<p>GitHub offers version control. This allows you to change portions of your code and test it before committing the changes to the main source file. This is done by &lsquo;branching&rsquo;&nbsp; &ndash; a copy of the main code is created to make changes. Once the changes have been finalized and tested, the branch undergoes &lsquo;merging&rsquo;. The changes made in the branch are implemented in the main source file. This process of version control allows programmers improve upon their existing code without losing any previous progress. GitHub also allows collaboration. Several people can simultaneously work together on a single set of source files. The following is a guide to navigating through GitHub.</p>

<p><a href="https://guides.github.com/activities/hello-world/">https://guides.github.com/activities/hello-world/</a>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;</p>

<p>GitHub could also be used in a career-forward capacity. Job recruiters and hiring managers often ask for a GitHub profile to assess one&rsquo;s coding abilities.&nbsp; Keeping an updated, neat GitHub profile of your projects and assignments would be a great opportunity to display your skillset. This is usually a very valuable asset to maintain!</p>

<p><img src="data:image/png;base64,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" style="float:left; height:211px; width:612px" /></p>