Yes, I let myself go and get sucked in by my origins. I just had to work on a database that is from my hometown (Barcelona) as a modest way to show my support to the fellow Catalans in these sad days (‘Stay strong and together’)

Why so many people do not show up for their medical appointments? Is it normal to have a 20% of noshows? Well, I though this was an interesting question that deserved, at least, my attention. I had to main reasons: the first one was the economic impact it has in the underemployment of public resources and, most importantly, the social cost that has in the patient, in his/her family and in the future.

This is a study trying to predict what are the most important factors to determine the cost of college. I know, I am not that original and, I know, it seems obviuos but… You never know!!! Also, I just wanted to do some scraping and enjoy something I used to hate when I was 5: SOUP. I had to do a serious work (at least for me) digging and digging from the college data API and I came out with some interesting outcome.

This is a study trying to predict the price of the real Estate in Ames IA (regression model) and the type of transaction that takes place normal or Abnormal (classification model). As a human being, you must admit that at some point in your life you thought you knew about something more than you actually did. Don’t be ashamed because, amongst other things, we walk around this world carrying a backpack full of assumptions. They are useful most of the times… but not always. So, let’s try to take some weight off this bag and figure out what are these things that really matter when establishing the price in the real estate. Ready to find out?

I am working in a Synthetic Database for Fraud Detection in banking transactions. This is another Kaggle project with over 6 million rows, being each row a bank transaction. One of the challenges is that this is a very unbalanced database with a very small amount of fraudulent transactions (less than 1%). Let’s start working.