Artificial intelligence (AI) Artificial intelligence (AI) is essentially the intelligence of machines or systems. It is a branch of computer science that involves creating machines or intelligent agents with human-like thinking and behavior. AI-enabled systems can reason, learn, and act autonomously even going beyond human capability. AI is a broad term that uses data (information)…
Logistic regression is a Machine Learning classification algorithm that is used to predict the probability of certain classes based on some dependent variables. There are many situations in the real world where we have binary outcomes. We will be interested in discrete output variables rather than continuous ones. Some of those situations consist of drug…
Classification is a supervised learning problem that assigns a class label, given input variables. The labels take on only a small number of discrete values.
In this article, I’ll explain everything about generalization and cross-validation techniques used in machine learning. Generalization The success of any machine learning model is in the ability to generalize well on previously unknown data. This means that if we train a machine learning model, we don’t just want it to learn to model the training…
Data is often messy. Many times, statisticians and data scientists will be dealing with missing data that can affect the accuracy and reliability of data analysis. In this article, we will be exploring the reasons for missing data and uncover different methods to handle missing data in Python. Reasons for missing data Missing data can…
Logarithms were one of the topics that I had to study in school and undergraduate classes. Even though I loved mathematics, I didn’t particularly enjoy learning logarithms. Maybe because I didn’t learn or understand it the correct way. Right from the beginning of my machine learning journey, I saw logarithms frequently applied in machine learning….
We have already discussed linear regression in some previous posts. Linear Regression is a linear model that maps a linear relationship between independent variable/s (input features) (x) and dependent variable (y). A common approach uses a least square cost function $J(\theta)$ to fit the linear regression model. In this post, we will discuss the probabilistic…
In many real-world scenarios, we may not have complete knowledge or assumptions about the underlying data distribution. If the relationship between the variables can be approximated by a straight line, we can use linear regression. But if the relationships are more complex or nonlinear, other techniques should be adopted. Consider the problem of predicting y…
There are certain things, you need to consider before splitting your data. We can do a little experimentation with the function train_test_split().
How do we ensure the model performs well on unseen data? The new data may not be available, but we can mimic the experience of having new data with a procedure called train-test-split
