ML User Side

Hi, today I am going to talk abut ML but not about the code. I am going to talk from the user side and how this cutting-edge technology can help to co-create with the user fit solutions to problems. If we developed user-centric products we should definitly consider using ML as a part of the final product in our solutions.

When you think in ML you have to think of the word "learn" and "useful". So What does it mean to learn something? You probably have your own definition of it but i would like to give you a definition that I considered, understanding the current state of development of AI technology and my knowledge of it. So in my opinion "learn" mean algorithms and digital programs that can learn useful things from the data and sensory input and apply the new knowledge to similar user problems.

Other important information you know about ML is that the machines do not understand anything about that they have learned. Those who understand what means the outputs or results generated from our ML model, after it learned , are the humans.

So what does it mean for a machine to learn something? It mean that a machine can take inputs, in differents ways or from different sources, and transform them to useful outputs to user. If the machine gave us this, We could say that the machine learned something.

But what is the challenge? Well one of the problems with the machine learning and AI in general is figuring out ho an algorithm can find the fit solution. I think to get the fit solution you need to start clearly defining two things, the outcomes and the outputs. Defining twi things as our goals for our machine learning project are keys in the whole process of building and using a machine-learning programs and all our effort will be worth.

Finally, I would like to share with you the main types of machine-learning algorithms:

• Supervised Learning. The key here is that you need to give the model the inputs and outputs.
• Unsupervised Learning. In this case the data just contains only inputs and the work of the algorithms look for structures and patterns in the data.
• Reinforcement learning. The software performs actions based on a acumulative reward. I encourage you to figure out about it. It is the future.

I invite you to know our service in?CONAUTI

or Contact me by?Whatsapp

Enrique Suárez

AI & Data Product Development Leader.

Source: Professional Data Engineer, Dan Sullivan, Google.

I see you in the next article.