Write some Software

Your assignment is to complete the k-Nearest-Neighbor based prediction on the Titanic data set. You need to turn in your code (whether it is in a .py Python file or a .ipynb Python notebook) and a write up where you document the things you did as part of this data analysis project. Think of the write-up as documentation that you would write for a team of fellow data scientists. It will detail information on some of the decisions you made in the analysis process as well as explain some of the major take-aways you were able to make as a result of the analysis. The write up should start with a short overview of the big problem you looked at, and then it should address each of the following graded items and briefly explain what the problem was, what you did to address it, and why you did it that way. In each case, make sure it is clear which graded item you're talking about so I don't accidentally miss it and not give you points for something you did. Data Cleaning (2 points) Fill in missing values (2 points) Convert everything used for prediction to numeric values (2 points) Remove unneeded columns (or develop a way to ignore it in the distance computation) (2 points) Normalize data using proportion-of-max method or Z-score method Making Predictions (5 points) Implement distance function Implement function to classify new instances via (5 points) k-Nearest-Neighbor (you can get 2 points if you only have 1-Nearest-Neighbor working, the remaining 3 for making it work with arbitrary k) (2 points) Weighted k-Nearest-Neighbor Evaluation (2 points) Properly split data into training and testing sets (5 points) Implement a function to measure accuracy (i.e., % correct) of predictions on the test set (5 points) Create a plot comparing weighted-vs.-unweighted versions of the algorithm for several different appropriate values of k (you'll want to include some very small k as well as some very large k). You must use matplotlib or some other Python-based plotting/visualization library for this. Make sure this plot is actually placed in the write-up - copy the image over to Word if you are using Word.