Getting started with Machine Learning

Hello people this is Akshay Khale a programmer and a machine learning enthusiast.

Machine Learning is there in the world since long time but now the topic is getting popular since the machines are getting better at decision making and are able to take decisions close to humans. Machines started to think like humans, machines started to see like humans(Machine Vision), machine started to understand human language(Natural Language Processing).

In fact machines can now do some tasks better than humans the best example can be Deep Blue (chess computer) developed by IBM known for being the first computer chess-playing system to win both a chess game and a chess match against a reigning world champion under regular time controls.

So being a Programmer, I was also thrilled with Machine Learning but I suck at maths and most of the tutorials, online training articles and videos are focused on Maths which were not easy for me to understand, I contacted many Machine Learning experts over the email and most of them replied with some blogs and suggested some books which were filled with wierd mathematical expressions.

After searching all over the internet I landed on machinelearningmastery.com by Dr. Jason Brownlee and I was back on track with Machine Learning though, I am not a Mathematician. After digging dipper in the Web I landed on springboard.com. I would recommend this website to anyone who is interested in a career in Machine Learning and Data Science. The Website provides different courses on Data Science right from Beginners level. They also provide Career Track with a well-crafted curriculum for a successful career in Machine Learning and Data science and If you are enthusiastic about machine learning and want to get started as quick as possible then I would like to recommend Free Machine Learning in Python course by Springboard.

So in this article I will be sharing how to get started with Machine Learning.

First step will be understaning machine learning concepts, concepts like Deep Learning, Data Science, Machine Learning, Artifical Intelligence and so on... for that you can start with a small book named Machine Learning for Absolute Beginners by Oliver Theobald.

Next step will be chosing the language, I started with Python because there are many different python packages for Machine Learning with Python, so If you are not familiar with python, I would recommend you to learn Python and installing Python packages with Pip which is a Python Package Manager.

The third step will be learning some python packages which will allow you to manage dataset, read/load remote dataset and data visualization. For that initially you can start with Numpy and Scipy for managing Datasets and you can start with Matplotlib for Data visualization and for loading Remote Data or loading data from files you can use Pandas. You don't get terrified by the names they are very simple to use and I am sure you will love these packages.

Since now you are comfortable with different packages now you can start with training models with some data and creating predictions on that data for that just head over to Kaggle and try to make sense of some some code. As a beginner I checked Titanic: Machine Learning from Disaster solution given by Omar El Gabry and also started some simple datasets like Iris Flower dataset example given here.

Next steps

Since now you are able to create predictions on models and are comfortable with Python Packages, you can now go for advanced machine learning libraries like Tensorflow, Theano and dive deep into machine learning.

Some useful links