Blog Post 10
Andrew Y -
Hello everybody! As you all know, this past week came with some pretty big changes and a lot of progressions for my project. I decided that it was time to stop data collection and start looking at the data in a way that we can actually draw conclusions from it. This is why this week, I’m actually going to be talking about data analysis. It started off pretty simple. I was able to create a line graph that compared the t-statistics of the results of each prediction algorithm to show how they varied from the real outcomes in respect to each other. Plotting this over an x-axis of multiple weeks was able to give more perspective about whether or not there was any actual variance. However, simply looking at the graph and making conclusions based on that would result in a guessing game, which is in no way scientific. In order to draw real results that would be fit for a scholarly research paper, this graph had to be paired with some statistical test to determine if these t-statistics were actually different from each other and, if they were, how they differed. So, we can think of this graph as a way to display this long process of data collection in an easy and digestible way for the audience, rather than a way of actually assessing the data. However, I am kind of stumped with how I want to approach this statistical test. Once that’s figured out, the hard part of this project will be complete, and I can get to work on framing it in the overall context. Thank you for tuning in and make sure to check back in next week.
- Andrew