Post 3: How Neural Nets Learn

The past few weeks have been intense. I’ve spent so many hours reading, watching, and doing whatever I can to learn more about neural nets. In my first post, I explained the basics of machine learning, and in my second post, I explained what a neural net actually looks like. Now it’s time to delve into what actually allows a neural net learn from its data.

[Read more…]

Post 2: Basic Structure of a Neural Net

In the past week, I’ve been reading and watching a lot about the way neural nets function. In addition, I’ve been trying to get my own neural net up and running. It’s been very challenging, but I found a new resource that’s been incredibly helpful: Andrew Ng’s Coursera series on machine learning. This course takes a slightly different approach from what I was using before – there is a heavy focus on learning somewhat modular algorithms that can be combined to create more robust models. I would strongly recommend the course to anyone with a good foundation in math and programming and an interest in predictive models. I didn’t expect this at the outset, but I have learned most quickly from this online course.

[Read more…]

Post 1: Artificial Neural Networks and a Machine Learning Primer

At the beginning of the summer, I was planning to work with sales data of videogames. My goals were to assemble my own data set using a number of sources and then to conduct some statistical analysis on this data set. Basically, I wanted to apply my knowledge of machine learning algorithms to see how well they performed on a problem that I found interesting. When I actually started working, though, I found that the methods that I could apply were the aspect of my Monroe project that I was interested in most. In particular, artificial neural networks grabbed my attention. I talked with my advisor a bit, and we agreed that it would be interesting to take this summer project in a new direction. Instead of focusing on making a data set and using a suite of statistical learning methods, I will be focusing on getting a better understanding of this exciting method.

[Read more…]

Modeling Sales of Interactive Media

This summer, I will be investigating an old hobby of mine: video games. Specifically, I want to create a model that predicts the total sales of any given video game one year after its launch. I will do this by assembling a data set containing video games and their predictors, the traits that have an effect on the variable being estimated. Some examples of these predictors include the year the video game is published, whether or not the game is a sequel, and the genre of the video game. After determining which predictors of video games are significant with respect to their sales, I will optimize my model, and I will test it for reliability in the real world.

[Read more…]