Start your Machine Learning Journey here with Python
Want to get started with Machine Learning (ML)?
It is a jungle out there with all the information available. How do you get started and what should you begin with?
On this page I will guide you through this jungle. The goal is to get you started with an practical approach. I do believe, the only way you will truly master something is by doing it yourself.
So let’s get started together on this journey.
Why learn Machine Learning?
Machine Learning has in many fields shown superior to the classical way to solve problems with computers.
Some of those areas include the following.
- Speech to text. Just think about it. How would you make a program which could understand an audio with speech and transform that to text. That seems like magic. Machine Learning is the key.
- Identify objects on images and videos. This is another amazing area. Identify objects, e.g, a bike, car, human, and so forth.
- Analyze sentiment in text. Is the text a positive or negative statement. This is important when analyzing a big amount of text. Say, a brand want to analyze all news posted about it. It would take forever to have someone read and analyze it. With machine learning this can be done fast and cheap.
- Prediction models. Predict what the next move? Predicting what next item a customer could be interested in buying. This is a billion dollar industry.
The list could go on.
Do you want to miss out on having skills in this area?
After following this list of tutorial, you will not master all af the above. But you will know the basis of Machine Learning.
- How does Machine Learning work compared to classical programming.
- How do you train your models. The overall process for having training data, test the accuracy of the model, and how to use the model.
- Main approaches, like supervised and unsupervised learning.
Prerequisites to follow this guide
What do need to know before you start this guide?
Good question. Not much, but it will greatly help you if you have the following.
- Some basic Python skills. See, you do not need to be a master in Python programming to use Machine Learning. All you need to understand to get started is the basics of a program structure.
- Time to actually do the programming yourself. Try all the examples along the way. Do not fool yourself that you understand it without trying. I have myself often just thought, “Oh, this is simple, I get it!”. To later realize that I did not master it well. Often details hide behind the simplicity.
- A healthy curiosity to try similar examples on your own. This will help you a lot of your understanding. First, read and follow along the tutorials. Second, make similar examples and see it work.
- Some simple math and statistics. Nothing advanced here. Just addition, subtraction, multiplication, division, simple average, maximum, minimum, and other simple concepts.
Yes, that is all you need.
Let’s get started!
Start simple. Start easy. Below you can follow the simple steps to get well on your way with Machine Learning using Python.
Follow all the steps in the given order.
Step 1: Get started
In the first tutorial, we will start by looking into the difference between classical computing and machine learning. It will continue to make a simple linear regression model with Python.
Linear regression is simple to understand and therefore a good starting point.
Step 2: My favorite: Reinforcement learning!
Yes, this is my favorite Machine Learning type. Reinforcement learning.
Why? Because it is the simples to understand. Also, you can program it all without using libraries.
This example uses simple Object Oriented approach to show you the power of it. But it is only used for the example it solves. The code for Reinforcement learning is done without any object oriented programming.
Step 3: Unsupervised Machine Learning
Another concept in Machine Learning is unsupervised machine learning, where you do not help the machine with categorizing the samples. You leave it all to the machine (algorithm).
Step 4: Play with Twitter data
The best place on earth to get real data from real people is on Twitter. No kidding. The data is available for free there, and it is a lot of fun to play with data from Twitter.
To get started with that you need to setup a Twitter developer account. It is all covered in the following tutorial.
Step 5: Use prediction model on Tweets
Using libraries out of the box is not straight forward. You need to know what you do.
Step 6: Sentiment model
One of the analysis often done in machine learning, how do people perceive the brand? Are the talking positively or negatively about it?
Here Twitter is amazing again. Collect tweets based on a subject/brand/word/hashtag and see the how a sentiment model determines whether people are in general positive or negative about it.
Step 7: Trading and Machine Learning
If you are interested in using a bot for trading. Well, first of all, I do not have the perfect answer. But try the following to get an idea on how to evaluate models.
Step 8: Use machine learning in data science
This is also a great place to use Machine Learning. Helping interpret data in a data science project.
This is only the start of the journey.
Feel free to ask me any questions, I will be happy to follow up on it.
Also, you can check out my courses.