These are chat archives for FreeCodeCamp/DataScience
discussion on how we can use statistical methods to measure and improve the efficacy of http://freeCodeCamp.com
Interpretis misspelled in the "two new problems" section. Otherwise, looks interesting! I've written my fair share of RMarkdown documents and this is a new one :smile: I'll also still need to get acquainted with D3 :laughing:
Very interesting today.
The first, insights about WaveNet, a neural network model to simulate the actual texture of raw audio. The sound of human voice was indistinguishable (think of Siri: you quickly get it is a robot...). What it is still not solved is assigning real language to that artificial voice though. We went through a code written in Keras: very nice!
The second was about the Panama papers problem, and analysing the data using networkX and community API. I also commented through this channel a previous meetup this summer where the same problem was analysed using neo4J.