"I want to learn machine learning, but I do not know what to start from where!"
I began listening to such voices frequently from beginners and beginners.
Along with AI and deep learning (deep learning),
"Machine learning", which has been getting hotter recently .However, even if you examine it, there are too many pieces of information, so you will not know where to start from.
And if you learn, I would like to learn as efficiently as possible without getting frustrated.
Therefore, this time, we will introduce five things that machine skill beginners absolutely recommend to beginners first to do!
If you follow the steps introduced here, you will be able to smoothly begin the introduction to machine learning !
1. Grab the overview first
Machine learning (machine learning) is one of the research subjects in artificial intelligence, and it is a research topic of engineering and technique that aims to realise the same function as the learning ability naturally performed by humans by computer .
Source: Wikipedia
Before going into learning "Machine Learning", first of all let's keep an overview of what it is!
- Features of machine learning
- Reasons to be noted
- Technology and method
2. Identify the purpose
In learning machine learning, input is inefficient if it is not fixed
what kind of purpose / purpose it is used for. There are five main uses for machine learning :
- Clustering : Based on the similarity of values, divide the given data into multiple groups. Used to group users' preferences.
- Classification : Assign the appropriate class to the given data. Used for categorising unsolicited e-mails, identifying images, etc.
- Recommendation : To guess information that is likely to respond from the user's past action history. Used for 'recommended' of EC site.
- Regression : To predict unknown values from past values. It is used for sales, stock price, device abnormality prediction and so on.
- Information Compression : Reduce the amount of data while maintaining the characteristics of data. Use it when leaving only one data when there is a correlated value.
3. Start learning with introductory content
Once you have an overview and purpose, let's get into the lesson!
In introductory learning of machine learning, I recommend that you start with simple slides, articles, learning sites, and so on.
When entering from a book etc., it is overwhelmed by difficult theory and words and it may be frustrated.If you like a slide, you can read it as crisp and still get rich information.
4. Deepen understanding with books
I think that we already have a lot of information, so you will be able to understand concepts and meanings of words.
Book recommendation for Machine Learning: meanwhile download this.
Book recomendation to learn Python : meanwhile download this too.
5. Actually move the hand
In addition to understanding with the head, it is the most familiar method to actually move hands and use it, so I'd like to use it somehow.
The convenience is "Python library" .Python has an excellent machine learning library so you only need to remember how to use it without knowing theory.
Summary
I would like to recommend it to beginners for machine learning First of all I have introduced five things to do.
Machine learning is difficult indeed, I think that there is an impression that is difficult to cling.
However, if you learn in the right way in the right order, you can master it firmly.
So I am pleased if this time's answer helps those, who are about to work on machine learning !
No comments:
Post a Comment