Second part of the TFG. In this part is developed a module capable of obtaining ROC curves for the evaluation of modified classifiers in the first part of the project.

This repository has the end result of the TFG carried out during 2016. The possibility of obtaining the results probabilistically rather than discrete results for further processing and obtaining ROC curves for evaluation are added to certain algorithms.

Detecting credit card fraud detection. Selecting an optimum threshold with analysis of confusion matrix and ROC curce

Survival modelling using Cox proportional hazard regression model

ROC Curve, Precision / Recall, Model Evaluation, Error Analysis. using MNIST dataset

simple script for plotting precision recall curves

This repository contains codes for running naive bayes and k-NN classification algorithms on large dataset in python

Making use of R programming, the analysis is focussed on the problem which insurance providers are facing today to define their target market and plan their sale strategies which helps them increase their market share and thereby, maximize their profitability. The analysis techniques used in the project are learnt through Data Analysis and Decis…

R Code and synthetic data accompanying the paper "Identifying patients using antidepressants for the treatment of depression"

This code analyzes a marketing campaign using Naive Bayes, Neural Network and Logistic

L2 Orthonormal Face Recognition Performance under L2 Regularization Term

Hyperspectral image Target Detection based on Sparse Representation

Random forest classifier model applied to a bank loan data to predict loan defaulters.

One Data Set with All Algorithms

This project used machine learning to understand characteristics of terrorist groups that engage in suicide bombings.

Machine learning models to classify news datasets as real or fake

Used the Global Terrorism Database to Explore Features of Suicide Bombings