Posts

Introdução Muitas pessoas querem aproveitar a praticidade dos modelos de linguagem para produzir conteúdo rápido e em grande escala, mas esse uso excessivo tem deixado claro que boa parte da internet já soa como escrita por máquinas. Mesmo que isso não surpreenda mais ninguém, devemos reconhecer que os modelos generativos estão aqui para ficar e precisamos aprender a usá-los com mais intenção e autenticidade. Leitores querem consumir conteúdos relevantes e originais na internet, mas o aumento de textos gerados por IA tem trazido materiais genéricos, marcados por padrões repetitivos que revelam sua origem artificial....

Types of Learning The types of learning determine what resources will be necessary to solve our problems. As suggested Mohri (2018), there are four fundamental types of learning, each with its specifications: Supervised machine learning Semi-supervised learning Unsupervised machine learning Reinforcement learning In this essay, we’ll use supervised machine learning for simplicity. Generally, we use this approach to handle regression problems and classification. A regression problem needs a real output, such as the mean value of a house, while classification typically results in a binary output (e....

Nowadays, even the most sophisticated tools that use AI are favored by the optimization algorithm, Gradient Descent. From ChatGPT to Midjourney, the concept is applicable due to the central theme in Machine Learning: minimizing the cost function. This post aims to give readers a general overview of this tool, which exists beyond the field of Machine Learning but plays a crucial role, being continuously mentioned in tutorials, books, or videos when the topic is Artificial Intelligence....

What is the utility of the perceptron? The perceptron was created in 1943 by Warren McCulloch and Walter Pitts. The fundamental idea behind its architecture is the ability to map an input value (x1, x2, …, xn) to a simple binary output value. However, the perceptron has some limitations, such as being useful only for binary classifications and requiring that the problem be linearly separable (which will be discussed later)....