What is machine learning
Since the beginning of time humans where good at certain things. They learned, read and acted on things. But because of the internet, the web has also developed a certain brain. They now can read, and render things we struggled when growing up. The machine doesn’t have the struggles of reading anymore, when we humans are start from scratch. The use of Artificial Intelligence(AI) are no secrets anymore, there are enough self driving cars, or scanning products. The computer is learning these things and it mostly recognize patterns. Not only that, the machine are also listening. Freaky right? Well how else are you auto generate Closed Captions on Youtube. It’s not the best at the moment, but it has come from far away and has some good things to it. The machine learns a lot by using big Data. Data is where everything comes from. It’s the key for Machine Learning(ML).
Machine Learning & Artificial Intelligence
AI systems are more close to human decisions. This includes certain things they can, problem solving, planning, learning, reasoning, knowledge representation, perception, motion and manipulation. However, ML is a data analytic technique where computer learns how to do natural things from humans or animals. It learns, or read a certain thing and show results from the data. It helps the user by giving insights from the data.
How does ML works?
ML has two types of techniques: Supervised learning, which trains a model on known input and output. Unsupervised learning which finds hidden patterns or intrinsic structure is input data
Supervised learning
The computer is exposed to big amount of data with labels, for example data about medicines. The computer needs to recognize the pixels and patterns from a picture and the create number or strings that later then can be used to search up in the database. With this human task are made easier.
Unsupervised learning
Using Unsupervised learning is mainly for identifying patterns in the data. It’s more categorized which spots similarities and cluster this together. Google uses this to group certain thing together like stories and similar topics. These patterns are hidden and would take a lot of time when trying to figure it yourself out. Or cellphone companies sorted by locations.
Machine Learning with basic knowledge of Javascript
Can I build something that needs ML without work experience of 10 years?
This research if for those who are interested in building a project that uses ML. The steps are quite intensive if you don’t know the basics of Javascript or working on a server. So if this is the case, work on your Javascript and start building with Nodejs.
But we are technies what do we need?
You need data. Look at your case, what do you want your machine to do? Does it need to analyze images or does it need to read text from a picture. There is a lot of use cases and you just need to find a library for Javascript.
-
A few I came across are:
- tesseractJS for reading text
- TensorFlow for basically everything from image to text.
Conclusion
This is just a short version about Machine Learning. There are tons to do research for but my main goal was get a small idea about this new era. Because I want to learn Frontend I rarely look at the new things or experiment with them. I’ve always thought that AI and ML are for die-hard coders. But with this little research I saw an opportunity to challenge myself in getting used to using new things.