An overview of Machine Learning and how business can apply it

This is the era of AI and machine learning. AI and machine learning are buzzwords for today’s technology and still growing at a rapid rate. However, most people are still confused about what it is and how it works. Therefore, this article is an attempt to explain what is machine learning and describe some types of machine learning methods.

1. What is machine learning?

This technology is the process of teaching a computer system how to make accurate predictions when fed data. It is a branch of Artificial intelligence or so-called AI. It originated from the idea that a system can learn from data, and patterns, and then replicate what has been taught.

2. Types of machine learning methods

There are 3 types of ML: supervised, unsupervised, and semi-supervised learning.

2.1. Supervised learning:

Supervised learning involves the learning process being conducted under the guidance of a supervisor or teacher and relies on ground truth for training. The correct answer is contained in the data itself, which is called “labeled” data. A supervised learning algorithm learns from these “labeled” data and then produces the desired outputs.

Supervised learning can be grouped into regression and classification problems:

– Classification: The output is classified into different categories such as “black” or “white”.

– Regression: A regression problem is when the output is a real value.

2.3. Unsupervised learning:

On the other hand, unsupervised learning does not have labeled outputs, allowing the model to work on its own to discover information, and deal with unlabeled data. There are two types of unsupervised machine learning: clustering and association.

– Clustering is the technique of attempting to divide data into groups clustered close together for instance the act of grouping customers by their purchasing behavior.

– Association This technique creates rules that explore the connections among data. For example, finding out if people buy this product will buy other similar products.

2.4. Semi-supervised learning:

The hybrid between supervised learning and unsupervised learning. Semi-supervised machine learning can utilize regression and classification models, as well as create predictions.

An example of semi-supervised learning is a photo archive where only some of the images are labeled while the majority are unlabeled.

3. Application of machine learning:

– Image recognition: one example of image recognition that you have already encountered on a daily basis is automatic friend tagging suggestions on Facebook. And the technology behind is machine learning’s recognition algorithm.

– Traffic prediction: Google Maps uses machine learning algorithms to show the shortest route and predicts traffic conditions in real time.

– Speech recognition: An example of this is the “search by voice” of Google. Nowadays, various applications of speech recognition with the usage of algorithms including Google Assistant, Siri, Cortana, and Alexa.

– Product recommendations: On some e-commerce sites, when buying a product you will see that they usually recommend similar products for you. Or when you search for a product in one place, it starts to show up on other websites while you are surfing. This is also one of the applications of machine learning algorithms.

 – Automatic language translation: Language is no longer the biggest barrier between people of different nationalities, as we now have a machine utilizing the algorithms of machine learning which could change any speech or convert any text to the language we know.

– Self-driving cars: You might not know this but self-driving cars are also one significant practical implication of machine learning.

– Virtual Personal Assistant: Nowadays, we have various personal assistants on different platforms such as Google Assistant, Alexa, Cortana, and Siri. These assistants work by recording our voice instructions, then send them over to the server on a cloud decode it using algorithms, and act accordingly.