Machine Learning in R
This project involved creating a machine learning model from any dataset that can be found online, using the R programming language and using R Sweave in particular to write a detailed report with every step explained and the findings. I used a customer churn dataset from a telecom company to product who will and who will not churn (i.e. leaving the company as a customer). For the model I used a simple white box decision tree because of its understandability and because it still performs really well (90+% accuracy) compared to a black box model such as SVM or neural networks. Evaluation of the models was done using 10-fold cross validation and by measuring accuracy and drawing the ROC curve/measuring AUC. It also includes visualizations to visualize the dataset that is being used. The images shown on the left are an excerpt of the report that was written and the code behind it. This is obviously not the entire code delivered for this project.
Rólam
Hi there! My name is Axel, I'm from Belgium and I am enthusiastic about anything to do with data, analytics, big data, artificial intelligence, machine learning and writing code (preferably in Python). My past projects on Freelancer.com usually had to do with - web scraping/automation - machine learning/data science/artificial intelligence/deep learning - sentiment analysis/natural language processing/text mining - optical character recognition (OCR) - big data topics/Apache Spark - SQL - general Python scripts to automate things - progressive web apps (Angular, Ionic, HTML5, CSS) If you need help in any of these areas, I will be more than happy to help!
