Machine learning models often operate in complex data environments where understanding the contribution of each feature to the model’s predictions is crucial. Determining feature importance is a key aspect of model interpretation, enabling us to grasp which factors significantly influence the model’s output. Let’s now explore different methods to determine Read more…
Outliers are data points that significantly differ from the rest of the data in a dataset. They are observations that lie at an abnormal distance from other values in a random sample from a population. Identifying and handling outliers is crucial because they can skew results and impact the performance Read more…
A normal distribution, also known as a Gaussian distribution, is a continuous probability distribution that is symmetrically shaped like a bell curve. The following set of statistical properties characterizes it: The normal distribution is a fundamental concept in statistics and probability theory, and it is widely used in various fields Read more…
Decision Trees are a fundamental model in machine learning used for both classification and regression tasks. They are structured like a tree, with each internal node representing a test on an attribute (decision nodes), branches representing outcomes of the test, and leaf nodes indicating class labels or continuous values. Decision Read more…
Ensemble Learning is a powerful method used in Machine Learning to improve model performance by combining multiple individual models. These individual models, also known as “base models” or “weak learners,” may have limitations such as high variance or high bias. There are two main types of ensemble models: There are Read more…
Setup In a previous post, we talked about the K-Means algorithm. In order to better understand how this algorithm works we will implement it from scratch in Python. To install the necessary tools please make sure to follow our setup posts either for Windows or for Linux. To improve the Read more…
Part II: Precision/Recall and Sensitivity/Specificity In the previous article on evaluation metrics for classification problems we explained that accuracy is a very useful metric and very easy to interpret. However, if the dataset we have is not balanced, it will be completely useless. We also defined what a confusion matrix Read more…
K-Means is an unsupervised Machine Learning algorithm used for data clustering. It is a technique employed to classify unlabelled data into a number K of clusters or groups based on their similarities. The number of clusters K is an input to the model. There exist different techniques to determine its Read more…
When training a Machine Learning model several scenarios can occur: The model can overfit, which means that the model is learning too much about the training data, it is too complex and it pays attention to very specific or random patterns within it or also called noise. This is a Read more…
Gradient Descent is an optimization algorithm for finding the local minimum of a function. It searches for the combination of parameters that minimizes the cost function. For the Gradient Descent to work, the cost function must be differentiable and convex. Visualization of Gradient Descent. Cost Function vs. Weight (W). Source: Read more…