In this machine learning fraud detection tutorial, I will elaborate how got I started on the Credit Card Fraud Detection competition on Kaggle. The goal of the task is to automatically identify fraudulent credit card transactions using Machine Learning. My Pythonic approach is explained step-by-step.

This spring, I taught a class on Machine Learning the Future at Cornell Tech covering a number of advanced topics in machine learning including online learning, joint (structured) prediction, active learning, contextual bandit learning, logarithmic time prediction, and parallel learning. Each of these classes was recorded from the laptop via Zoom and I just uploaded the recordings to Youtube.

In this tutorial, you will learn how to find a combination of stocks with high expected return and low risk using Python.

In their book “Perceptrons” (1969), Minsky and Papert demonstrate that a simplified version of Rosenblatt’s perceptron can not perform certain natural binary classification tasks, unless it uses an unmanageably large number of input predicates. It is easy to show that with sufficiently many input predicates, a perceptron (even on this type) can perform any classification with perfect accuracy (see page 3 of the notes below). The contribution of Minsky and Papert is to show that meaningful restrictions on the type of input predicates hamper the expressive ability of the perceptron to such a degree that it is unable to e.g. distinguish connected from disconnected figures, or classify according to whether the number of active pixels is odd or even. The former has a simple picture proof, whereas the crucial ingredient for the latter is the action of a permutation group on the retina (i.e. the input array) of the perceptron.

Kaggle‘s Quora Question Pairs competition has just closed, I’m pleased to say that with 10 days effort I ranked in the top 39th percentile (rank 1346 of 3396 in the private leaderboard). Having just run and spoken at PyDataLondon 2017, taught ML in Romania and worked on several client projects I only freed up time right at the end of this competition. Despite joining at the end I had immense fun – this was my first ‘proper’ Kaggle competition.

I have taken my own advice and started playing a few games in Japanese to try to learn the language better.  I record them and then upload to YouTube.  Here are a few of my current playlists.

The previous post on this blog sought to expose the statistical underpinnings of several machine learning models you know and love. Therein, we made the analogy of a swimming pool: you start on the surface — you know what these models do and how to use them for fun and profit — dive to the bottom — you deconstruct these models into their elementary assumptions and intentions — then finally, work your way back to the surface — reconstructing their functional forms, optimization exigencies and loss functions one step at a time.

For a while now I’ve been experimenting with ways to use generative neural nets to make portraits. Early experiments were based on deepdream-like approaches using backprop to the image but lately I’ve focused on GANs. As always resolution and fine detail is really difficult to achieve. For starters the receptive field of thse networks is usually less than 256x256 pixels. One way around this is tiling combined with stacking GANs, which many people have experimented with, for example this paper uses a two-stage GAN to get high resolution: (https://arxiv.org/abs/1612.03242).

TLDR: What if it was possible for surveillance to only invade the privacy of criminals or terrorists, leaving the innocent unsurveilled? This post proposes a way with a prototype in Python.

COLT 2017 accepted papers