To test the behavior of an algorithm and determine if it works properly, I created a test set with 3 dimensions and 3 classifications from distributions that are defined. In this case they are represented by 3 separate spheres in different quadrants of space. I generated 5% errors into the test set. For example,… Continue reading Random Forest Machine Learning Modeling
Category: Data Science
Rob’s Estimate of a Parabolic Air Conditioner BTU Sizing function
I whittled down the report to focus on what matters to me, which is simply the cooling power, or BTUs. SEER ratings are also important, to save electricity, but what is the point of efficiency if there isn’t enough cooling power?
When to Sell, Revisited
Part III. Introduction To understand the effects of a strategic rebalancing schedule, I began a study that compares multiple rebalancing strategies. The premise is that people cannot decide when to sell, and systematic methods to buy and sell would take emotions out of the equation and might work better. I choose to start with a… Continue reading When to Sell, Revisited
To Machine Learn or not to machine learn?
To me the difficulty in processing machine learning is that it provides no straightforward formulaic relationship at all; it’s a black box.