Ahhh what a nerdy evening. I decided it would be interesting to apply some regression algorithms to help predict when the James Webb Space Telescope would ACTUALLY be finished based on past NASA estimates. Some background: The James Webb was originally estimated in 1997 to be completed in 10 years (2007) for a cost of $0.5 Billion. As of last year (2013), the James Webb is predicted to be completed in 5 more years (2018). We could be mean and predict that every 16 years, the James Webb telescope is another 50% closer to being completed, but that would put the telescope at 95% done in 160 yrs total. Ahh converging geometric series. Brings back memories of Krista Taylor Maxsons Calc 3. Luckily, I think we can assume they will complete the project at some time. Now, if we instead apply a simple linear regression, it looks like it will be finished in 2032 for a total cost of $18.4 billion. Alternatively, we can do a second power polynomial regression where the estimated completion date is 2026 for a cost of $15.25 billion (the cost calculated linearly). My hope is that they can actually meet their goal of 2018 for a cost of $8.8 billion, but I dont hold out a lot of hope. PS: r > .9 for all three of those regressions. So, to all you students of Dr. Doug Darbro, remember, when you get out in the world, you, too, can sit bored in the evenings and make dire predictions about the future of our space program! Alternatively, you may be using your super regression skillz to model lens distortion in fisheye lenses for software-based correction. Thats a bit more fulfilling.
Posted on: Fri, 19 Sep 2014 02:48:21 +0000
Trending Topics
Recently Viewed Topics
© 2015