Often when talking about Machine learning or statistics, there is a perception that people can solve all the world's problems with magic boxes called algorithms that no one really knows how they work.
This collection of articles does not aim to explain all the knowledge of statistics and Machine learning, nor to explain the main topics that make up this vast world, but simply encompasses some topics that I consider important and interesting. Some of these will be more discursive, some more computational, and others more mathematical-statistical, but each article will try to maintain an applied approach to make the explanation clearer.
Articles
This article covers the basic methods of random and pseudo-random number generation.
Article link
From classical statistics to Machine Learning evolution of data.
Article link
Stationarity and non-stationarity of a time series
Article link
Multivariate stochastic processes: VARMA and VAR models
Article link
Cointegration
Article link
This article covers the Monte Carlo Method, also analyzing variance reduction methods such as: Control variables method and antithetic variables method.
Article link
What is data science.
Article link
Stacked models.
Article link
Fisher's exact test is a hypothesis testing test used in non-parametric statistics in situations with two dichotomous nominal variables.
Article link
The measures underlying Network analysis in R
Article link
The measures underlying Network analysis in Python
Article link
Download dataset from_to.csv
Download dataset info.csv
Work in progress: Dynamic network representation
Work in progress: Cluster and community analysis
Forecasting theory
Article link
Convolutional neural networks: definition and elements.
Article link
Correlation vs causality
Article link
Schematic notions on the steps in market research
Article link
The multiple testing problem.
Article link
Four basic components of these models: Trend, Cycle, Seasonality, Noise (WN).
Also included a small insight on the use of regressors in these models.
Article link
Boosting is a machine learning technique that falls into the ensemble learning category. The idea of Boosting is to combine "weak" classifiers in order to create a classifier with better accuracy.
Article link
This article explains in an introductory way the concept of neural network and introduces the Multi-Layer Perceptron (MLP)
Article link
It is a non-parametric regression technique and can be seen as an extension of linear models that automatically models non-linearities and interactions between variables
Article link
Record Linkage and Knowledge Discovery
Article link
Introduction to the Bayesian approach
Article link
Prior selection:
a) Direct assignment
b) Conjugate distributions to the Model
c) Non-informative distributions
Article link
Hyperparameter selection methods
Article link
Predictive inference with the Bayesian approach:
a) Point estimation
b) Interval estimation
c) Hypothesis testing
Article link
Posterior synthesis
Article link
1. Introduction to State Space models
Article link
2. Introduction to UCM models in State Space form with the KFAS library
Article link
3. State Space models estimation of unknown parameters
Article link
4. Application of State Space models estimation of unknown parameters and auxiliary residuals
Article link
5. Forecasting, filtering and smoothing
Article link
Summary of the steps of an analysis, from data to prediction.
Article link
Psychometrics and the Myers Briggs personality test as used in market analysis
Article link
0. Introduction to geostatistical data.
Article link
1. Introduction to spatial statistics
Article link
2. Spatial data in R
Article link
3. Spatial Point Processes: The Poisson process
Article link
4. Spatial Point Processes: The test for CSR
Article link
5. Spatial Point Processes: Estimation of the intensity function
Article link
6. Introduction to Geostatistics, large and small scale variability.
Article link
7. Geostatistical Model.
Article link
8. Exploratory analysis EDA and ESDA.
Article link
9. Theoretical introduction to spatial prediction and kriging and application in R on Ordinary and Universal Kriging.
Article link
Text
Article link
Text
Article link
Text
Article link
Text
Article link
Text
Article link
Text
Article link
Text
Article link
Text
Article link
Text
Article link
Text
Article link
Text
Article link
Text
Article link
Text
Article link
Text
Article link
Text
Article link
Text
Article link
Text
Article link