Category Statistics

Visual manipulation of data Visual manipulation of data

Sin categoría, Statistics

The megapixel trap

Visual manipulation of data using poorly designed charts can distort data interpretation. The most common errors, such as missing axes, manipulated scales, and confusing pie charts, are described, which can lead to erroneous conclusions. Learning to detect these errors will allow us to improve our ability to visually analyze and interpret data.

Manuel Molina
02/04/2025

Machine learning, Sin categoría, Statistics

Apophenia

Overfitting occurs when an algorithm over-learns the details of the training data, capturing not only the essence of the relationship between them, but also the random noise that will always be present. This negatively affects its performance and its ability to generalize when we introduce new data, not seen during training.

Manuel Molina
11/27/2024

Machine learning, Statistics

The wisdom of the weirdwoods

Simple decision trees have the problem of being less accurate than other regression or classification algorithms, as well as being less robust to small modifications of the data with which they are built. Some techniques for building ensemble decision trees are described, such as resampling aggregation (bagging) and random forests, which aim to improve the accuracy of predictions and avoid overfitting of models.

Manuel Molina
11/26/2024

Sin categoría, Statistics

The Alchemist

The Egger’s test is the most popular quantitative method to assess funnel plot asymmetry. It is based on a linear regression model between the effect measurement and the precision of the studies. A non-zero intercept value indicates asymmetry in the funnel plot probably due to a probable publication bias.

Manuel Molina
11/05/2024

Sin categoría, Statistics

A seance

The existence of publication bias can alter the results of a meta-analysis. The trim and fill method attempts to calculate an estimate of the effect corrected for bias that may have been introduced by missing studies. The objective is to impute these missing studies and include them in the funnel plot until the asymmetry disappears. Once this extended funnel is achieved, the effect measure is recalculated to obtain an estimate that corrects the effect of small studies.

Manuel Molina
09/10/2024