Project requirements (professor):
– Pick a reasonably complex data set. Talk to me if unsure whether this is the case for your data.
– Import the data into Pandas and clean/wrangle/transform it.
– Perform a basic exploratory analysis and write up a summary
– Separate the data into a training set and a test set and create a basic prediction model, for example, using linear regression.
– Verify the validity of the predictor and summarize your results
– Select a more advanced prediction method and describe how the method works step-by-step
– If the method requires tuning, select a set of parameters until the error on the test set is minimal
Project details:
Deadline: 28.11.2017 23:30 UTC time
Data: Stock and Cryptocurrency data from 01.01.2014 to 10.31.2017.
It is required that a meaningful statistical analysis be preformed on the data that leads to a prediction model done in Python3. Included file is a Python notebook with the data loaded and initial correlation matrix done, what is required is that simulations be performed and data analysed correctly, see below.
Assignment:
• preform monte carlo prediction regarding the stock prices (percentage change of prices) after using a Copula distribution (spearman correlation)
• perform monte carlo prediction regarding the stock prices (percentage change of prices) using a multivariate normal distribution
• co-integration
• portfolio suggestion based on analysis
