A starting guide to artificial intelligence
If you are a software developer or even not in the software field and want to start a career in Artificial intelligence, then this is the right article for you!
As a Deep Learning (Artificial intelligence) Engineer, I faced many problems with starting with the field, especially since there are many resources and you may get lost in all of them. Here, I will be writing a plan for you where you can start in a clear and well-organized way.
When I started reading about artificial intelligence I felt like this is a huge field and it’s so difficult to get into it. In fact, It is a huge field but you have to know the correct approach to get into it. What comes to your mind when we say artificial intelligence is some robots trying to attack people’s life and increase unemployment. At least this is what came to my mind at the beginning. However, I faced a lot of challenges to start learning this field. A lot of times I felt completely lost. I remember when I had 0 experience in Artificial intelligence, I started learning Convolution neural networks which was a nightmare as I started immediately with one of the advancements of the field without learning the basics. you may ask yourself, What is deep learning? Should I start learning deep learning first? How can I make a machine-learning model?
First of all, Artificial intelligence simply is defined as the way that makes machines think and process information like the human brain. Not like traditional programming.
Artificial intelligence is a very huge field and you can’t say that you learn artificial intelligence. Many definitions come under artificial intelligence such as Machine learning, Deep Learning, Unsupervised learning, and Reinforcement learning. But what are all these exactly?
To make it more clear, Machine learning is simply teaching the machine and making the machine works like a human brain. It is divided into 3 parts. Supervised learning, Non-supervised learning, and reinforcement learning. Supervised learning is simply teaching the machine to predict something and you teach the machine by comparing the predicted result with the real result. This process is called training. So you know the real result. However non-supervised learning is teaching the machine without giving it the real result. So the machine relies on collecting data that has a similarity. Finally, reinforcement learning is letting the machine learn on its own by trying different actions, then the machine will get rewarded or penalized depending on how the action affected the environment.
So the question now is how to start learning machine learning? and what exactly do you need?
Fortunately, machine learning is a field that you can work in without obtaining a degree in computer science/engineering. You will need at the beginning a basic knowlegde of Calculus and Linear algebra. Very basic knowledge will be enough you do not have to be an expert! and if you don’t have good coding experience it's fine, this guide will help you with this!
Then I would strongly recommend studying the free course Machine learning. The course is taught on Coursera by professor Andrew NJ. This course is focusing on the math behind machine learning and implementing it in MATLAB. I would strongly recommend this course. It is the best way to start your journey.
Now after you have finished this course. You should have strong knowledge of the math behind machine learning. This will help you a lot in the future to deploy and optimize machine learning problems. Now is the time for learning the most popular programming languages that are used in machine learning. Yes, it is not MATLAB. Python and R are the most popular languages. The reason is that they are free and open source. The huge amount of documentation and ease of syntax makes them very popular. Python is used to analyze data, build ML models, and a lot more. R is used for statistical analysis.
Don’t worry it will not take a lot of time. You just need 4 hours to finish this video. Python for beginners by freecodecamp. After this video, you will feel a lot of comfortable writing a complete code by Python. However, it is 2 hours only to learn the R language. Using also Full R course by freecodecamp.
Now you are having enough knowledge for machine learning from both perspectives (Mathematical and Coding).
Now you need one more hour to watch this one, it is combining what you have learned about Python and Machine learning together and building a complete model.
After you build your first model. you should now try to build more models and upload them to GitHub. Now you are an expert in Machine learning.
In part 2, we will see the most important python frameworks that will be used in data manipulation and building basic machine learning models.
