This section shows some of the projects I have done as a result of my learning process or hobby. All these exercises are for purely educational purposes.
Machine Learning
-
Pandemic Board Game
An alternative implementation of Pandemic Board Game through the develop of artificial intelligence agents who model the standard play of an average user.
-
Multi-agent system examples
Using Jason framework for the development of multi-agent systems.
-
Fish bank simulation
Using Java to model a 2D simulation of a fish bank autonomous behaviour environment.
-
Agents Communication
Using Jade to develop a simple comunication between autonomous agents.
-
Artificial Inmune System (AIS) Development
Implementing an artificial immune system (AIS) to solve the N Queen Problem.
-
K-nearest neighbors algorithm to predict a car’s market price
In this project, we’ll use some of the machine learning techniques to predict a car’s market price using its attributes. The chosen technique will be k-nearest neighbors algorithm developed in its univariate and multivariate models. Afterwards, another part related to the optimization of the best model obtained is executed. To do this, we’ll make use of grid search techniques.
-
Linear regression model to predict House Sale Prices in the market
For this project, we’ll see how to develop the machine learning approach with a linear regression model for predicting house sale prices at the same time as we understand different techniques of feature selection and feature engineering to accomplish a good model fitting. Some of these techniques include cleaning, transforming, and selecting appropiate features.
-
Linear regression model to predict future prices in the stock market
Inside this project a simple linear regression model is developed to extract information in the stock market and use it to predict next prices in it.
-
Linear regression, Decision Tree and Random Forest models to predict bike renting in a communal bike sharing station
For this project, we’ll see how to develop a linear regression model, a decision tree model and a random forest model to predict the number of people renting bikes at a communal bike sharing station located in Washington D.C., US. We’ll evaluate every model and define what it is the one that best fits the data and gets the best overall results.
Deep Learning
-
Handwritten Digits Classifier
Comparison of a k-nearest neighbors (KNN) algorithm and different architectures of a multilayer perceptron (MLP) feedforward neural network model for an image classification problem on handwritten digits.
-
Clouds Classifier
Using Deep Neural Networks to build different cloud classifier architectures (with Generative Adversarial Networks for Data Augmentation).
-
CIFAR10 Classifier
Building different Deep Neural Networks architectures to accomplish good classification metrics for Cifar10 dataset.